AI Proxy · benchmark
2026-08-05 06:42 · proxy 0.2.0
DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4 leads: 89% fully correct at 13.3 tok/s out, 2 s to load.
5 other configurations scored the same and were slower. Ranked by correctness first, then output rate. 96 configurations measured across 3 backends.
Run this — 70/30 weighted
gemma4:26b
87% correct · 63.2 tok/s · 16.8 GB · answers in 3.0s · score 70
Runner-up
ornith-nvfp4
85% correct · 65.6 tok/s · answers in 3.1s · score 69
Don't be fooled by
qwen3:0.6b
295 tok/s and only 3% correct
Every configuration, placed by correctness and output rate. A point below and to the left of another is beaten on both counts at once, so the dashed frontier is the shortlist — everything off it is dominated by something on it. Hover any point for its name.
Think
off
Temp
0.0
TTFT is the first token of any kind; TTFC the first content token — the gap between them is time the model spent reasoning. Decode rate is measured from the first token onward, so reasoning tokens count as generated work. Best value in each column is highlighted.
| Configuration | Load | Resident | TTFT p50 | Decode p50 | Tokens | Total p50 | Fully correct | Cases | vs best | OK |
|---|---|---|---|---|---|---|---|---|---|---|
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4 | 2 s | — | 1,389 ms | 13.3 | 184 | 11,693 ms | 89% | 90% | 23.9x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4 | 2 s | 0.2 GB | 1,419 ms | 13.3 | 185 | 11,993 ms | 89% | 90% | 24.5x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4 | 61 s | 0.4 GB | 1,489 ms | 13.2 | 186 | 13,272 ms | 89% | 90% | 27.2x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4 | 60 s | 0.4 GB | 1,433 ms | 13.2 | 186 | 11,691 ms | 89% | 90% | 23.9x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4 | 2 s | 0.2 GB | 1,461 ms | 13.2 | 185 | 11,748 ms | 89% | 90% | 24.0x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4 | 61 s | 0.5 GB | 1,426 ms | 12.3 | 186 | 12,970 ms | 89% | 90% | 26.5x | 94/94 |
| gemma4:26b · ollama · cached · 1 | 17 s | 17.9 GB | 474 ms | 63.2 | 265 | 2,991 ms | 87% | 90% | 6.1x | 94/94 |
| gemma4:26b · ollama · cached · 4 | 17 s | 17.9 GB | 833 ms | 36.6 | 244 | 5,848 ms | 87% | 91% | 12.0x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1 | 60 s | 0.0 GB | 356 ms | 17.6 | 197 | 9,318 ms | 87% | 90% | 19.1x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1 | 60 s | 0.4 GB | 341 ms | 17.6 | 197 | 9,299 ms | 87% | 90% | 19.0x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1 | 2 s | — | 341 ms | 17.6 | 197 | 9,310 ms | 87% | 90% | 19.1x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1 | 2 s | — | 363 ms | 17.6 | 197 | 9,340 ms | 87% | 90% | 19.1x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1 | 60 s | 0.1 GB | 338 ms | 17.5 | 197 | 9,354 ms | 87% | 90% | 19.1x | 94/94 |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1 | 2 s | 0.3 GB | 340 ms | 17.5 | 197 | 9,344 ms | 87% | 90% | 19.1x | 94/94 |
| gemma4:26b · 16.8 GB · ollama · cold · 1 | 20 s | 17.9 GB | 520 ms | 63.2 | 268 | 3,038 ms | 86% | 90% | 6.2x | 94/94 |
| ornith-nvfp4 · NVFP4 · vllm · cold · 1 | 0 s | 0.1 GB | 118 ms | 65.6 | 231 | 3,074 ms | 85% | 93% | 6.3x | 94/94 |
| gemma4:26b · ollama · cold · 4 | 1 s | 17.9 GB | 829 ms | 37.7 | 241 | 5,059 ms | 85% | 91% | 10.4x | 94/94 |
| qwen3-coder-next · NVFP4 · vllm · cold · 1 | 1 s | 0.1 GB | 160 ms | 62.1 | 222 | 3,259 ms | 84% | 87% | 6.7x | 94/94 |
| qwen3-coder-next · NVFP4 · vllm · cached · 1 | 18 s | — | 152 ms | 61.8 | 202 | 2,942 ms | 84% | 87% | 6.0x | 94/94 |
| qwen3-coder-next · NVFP4 · vllm · cached · 4 | 17 s | — | 424 ms | 42.3 | 206 | 4,739 ms | 84% | 89% | 9.7x | 94/94 |
| ornith-nvfp4 · NVFP4 · vllm · cached · 4 | 13 s | — | 324 ms | 49.2 | 235 | 3,951 ms | 82% | 88% | 8.1x | 94/94 |
| gpt-oss:120b · 60.9 GB · ollama · cold · 1 | 17 s | 60.4 GB | 597 ms | 42.4 | 602 | 13,706 ms | 82% | 81% | 28.0x | 94/94 |
| gpt-oss:120b · ollama · cached · 1 | 25 s | 60.4 GB | 505 ms | 42.4 | 602 | 13,828 ms | 82% | 81% | 28.3x | 94/94 |
| ornith-nvfp4 · NVFP4 · vllm · cached · 1 | 16 s | — | 117 ms | 65.4 | 245 | 3,120 ms | 81% | 91% | 6.4x | 94/94 |
| ornith-nvfp4 · NVFP4 · vllm · cold · 4 | 0 s | — | 325 ms | 48.6 | 257 | 3,993 ms | 81% | 88% | 8.2x | 94/94 |
| qwen3-coder-next · NVFP4 · vllm · cold · 4 | 0 s | — | 405 ms | 42.3 | 204 | 4,529 ms | 81% | 86% | 9.3x | 94/94 |
| qwen3.6:35b-a3b · ollama · cached · 4 | 14 s | 22.0 GB | 4,847 ms | 74.9 | 295 | 7,759 ms | 79% | 88% | 15.9x | 94/94 |
| qwen3.6:35b-a3b · ollama · cold · 4 | 1 s | 22.0 GB | 4,687 ms | 74.8 | 295 | 7,831 ms | 79% | 88% | 16.0x | 94/94 |
| qwen3.6:35b-a3b · ollama · cached · 1 | 14 s | 22.0 GB | 410 ms | 74.8 | 295 | 4,168 ms | 79% | 88% | 8.5x | 94/94 |
| qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1 | 10 s | 22.0 GB | 423 ms | 74.8 | 295 | 4,188 ms | 79% | 88% | 8.6x | 94/94 |
| qwen3.6:27b · 16.2 GB · ollama · cold · 1 | 10 s | 16.5 GB | 608 ms | 12.1 | 265 | 16,236 ms | 79% | 80% | 33.2x | 94/94 |
| qwen3.6:27b · ollama · cached · 4 | 86 s | 16.5 GB | 21,008 ms | 12.0 | 265 | 36,906 ms | 79% | 80% | 75.5x | 94/94 |
| qwen3.6:27b · ollama · cold · 4 | 2 s | 16.5 GB | 23,058 ms | 12.0 | 265 | 40,354 ms | 79% | 80% | 82.6x | 94/94 |
| qwen3.6:27b · ollama · cached · 1 | 86 s | 16.5 GB | 619 ms | 12.0 | 265 | 16,259 ms | 79% | 80% | 33.3x | 94/94 |
| gpt-oss:120b · ollama · cold · 4 | 1 s | 60.4 GB | 1,327 ms | 19.8 | 624 | 30,143 ms | 77% | 76% | 61.7x | 94/94 |
| devstral-2:123b · 69.8 GB · ollama · cold · 1 | 25 s | 92.9 GB | 740 ms | 2.6 | 174 | 52,658 ms | 77% | 84% | 107.8x | 94/94 |
| devstral-2:123b · ollama · cached · 4 | 265 s | 92.9 GB | 2,885 ms | 2.5 | 171 | 63,463 ms | 77% | 85% | 129.9x | 94/94 |
| devstral-small-2:24b · 14.1 GB · ollama · cold · 1 | 11 s | 24.7 GB | 348 ms | 13.7 | 175 | 10,686 ms | 76% | 86% | 21.9x | 94/94 |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4 | 1 s | 20.7 GB | 4,333 ms | 63.3 | 243 | 7,093 ms | 74% | 81% | 14.5x | 94/94 |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1 | 12 s | 20.7 GB | 406 ms | 63.3 | 243 | 3,206 ms | 74% | 81% | 6.6x | 94/94 |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4 | 17 s | 20.7 GB | 4,040 ms | 63.3 | 243 | 7,698 ms | 74% | 81% | 15.8x | 94/94 |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1 | 17 s | 20.7 GB | 399 ms | 63.3 | 243 | 3,190 ms | 74% | 81% | 6.5x | 94/94 |
| laguna-xs-2.1 · ollama · cold · 4 | 0 s | 20.2 GB | 502 ms | 52.7 | 251 | 4,524 ms | 74% | 86% | 9.3x | 94/94 |
| gpt-oss:120b · ollama · cached · 4 | 25 s | 60.4 GB | 1,277 ms | 22.1 | 629 | 28,981 ms | 74% | 74% | 59.3x | 94/94 |
| devstral-2:123b · ollama · cached · 1 | 265 s | 92.9 GB | 692 ms | 2.6 | 172 | 52,673 ms | 74% | 84% | 107.8x | 94/94 |
| devstral-2:123b · ollama · cold · 4 | 7 s | 92.9 GB | 2,696 ms | 2.5 | 171 | 60,367 ms | 74% | 85% | 123.5x | 94/94 |
| laguna-xs-2.1 · ollama · cached · 4 | 12 s | 20.2 GB | 518 ms | 51.0 | 267 | 4,037 ms | 73% | 83% | 8.3x | 94/94 |
| qwen3-coder:tuned · ollama · cold · 4 | 1 s | 50.6 GB | 4,216 ms | 60.7 | 205 | 7,419 ms | 72% | 83% | 15.2x | 94/94 |
| qwen3-coder:tuned · 48.2 GB · ollama · cold · 1 | 11 s | 50.6 GB | 339 ms | 60.7 | 205 | 3,091 ms | 72% | 83% | 6.3x | 94/94 |
| qwen3-coder:tuned · ollama · cached · 1 | 17 s | 50.6 GB | 367 ms | 60.7 | 205 | 3,112 ms | 72% | 83% | 6.4x | 94/94 |
| qwen3-coder:tuned · ollama · cached · 4 | 17 s | 50.6 GB | 4,080 ms | 60.7 | 205 | 7,255 ms | 72% | 83% | 14.8x | 94/94 |
| qwen3-coder-next · ollama · cached · 4 | 17 s | 48.7 GB | 4,936 ms | 60.1 | 214 | 7,620 ms | 72% | 82% | 15.6x | 94/94 |
| qwen3-coder-next · 48.2 GB · ollama · cold · 1 | 11 s | 48.7 GB | 380 ms | 60.1 | 214 | 3,243 ms | 72% | 82% | 6.6x | 94/94 |
| qwen3-coder-next · ollama · cold · 4 | 1 s | 48.7 GB | 4,258 ms | 60.1 | 214 | 7,746 ms | 72% | 82% | 15.9x | 94/94 |
| qwen3-coder-next · ollama · cached · 1 | 17 s | 48.7 GB | 390 ms | 60.1 | 214 | 3,229 ms | 72% | 82% | 6.6x | 94/94 |
| devstral-small-2:24b · ollama · cached · 4 | 66 s | 24.7 GB | 846 ms | 12.5 | 173 | 12,638 ms | 72% | 85% | 25.9x | 94/94 |
| devstral-small-2:24b · ollama · cold · 4 | 1 s | 24.7 GB | 892 ms | 12.4 | 175 | 13,571 ms | 72% | 85% | 27.8x | 94/94 |
| gemma3:27b · ollama · cached · 1 | 91 s | 17.9 GB | 632 ms | 11.5 | 227 | 16,793 ms | 72% | 79% | 34.4x | 94/94 |
| gemma3:27b · 16.2 GB · ollama · cold · 1 | 10 s | 17.9 GB | 655 ms | 11.5 | 225 | 16,659 ms | 72% | 79% | 34.1x | 94/94 |
| gemma3:27b · ollama · cold · 4 | 2 s | 17.9 GB | 1,163 ms | 10.3 | 228 | 20,135 ms | 72% | 80% | 41.2x | 94/94 |
| devstral-small-2:24b · ollama · cached · 1 | 73 s | 24.7 GB | 340 ms | 13.7 | 170 | 10,670 ms | 71% | 85% | 21.8x | 94/94 |
| gemma3:27b · ollama · cached · 4 | 91 s | 17.9 GB | 1,201 ms | 10.2 | 228 | 22,711 ms | 71% | 79% | 46.5x | 94/94 |
| qwen3-coder:30b · 17.3 GB · ollama · cold · 1 | 7 s | 24.0 GB | 247 ms | 83.7 | 180 | 1,974 ms | 70% | 86% | 4.0x | 94/94 |
| qwen3-coder:30b · ollama · cold · 4 | 0 s | 24.0 GB | 438 ms | 42.3 | 180 | 4,063 ms | 69% | 87% | 8.3x | 94/94 |
| laguna-xs-2.1 · 18.9 GB · ollama · cold · 1 | 7 s | 20.2 GB | 261 ms | 86.6 | 239 | 2,215 ms | 67% | 79% | 4.5x | 94/94 |
| qwen3-coder:30b · ollama · cached · 1 | 13 s | 24.0 GB | 216 ms | 83.6 | 176 | 1,919 ms | 67% | 86% | 3.9x | 94/94 |
| qwen3-coder:30b · ollama · cached · 4 | 13 s | 24.0 GB | 430 ms | 41.3 | 184 | 4,169 ms | 67% | 86% | 8.5x | 94/94 |
| laguna-xs-2.1 · ollama · cached · 1 | 12 s | 20.2 GB | 184 ms | 86.6 | 246 | 2,195 ms | 66% | 79% | 4.5x | 94/94 |
| gemma4 · ollama · cached · 4 | 19 s | 4.0 GB | 742 ms | 42.4 | 483 | 10,835 ms | 65% | 71% | 22.2x | 94/94 |
| gemma4 · ollama · cached · 1 | 19 s | 4.0 GB | 470 ms | 55.4 | 491 | 7,721 ms | 64% | 67% | 15.8x | 94/94 |
| gemma4 · 8.9 GB · ollama · cold · 1 | 12 s | 4.0 GB | 479 ms | 55.4 | 491 | 7,729 ms | 64% | 67% | 15.8x | 94/94 |
| llama4 · ollama · cached · 4 | 47 s | 64.0 GB | 2,943 ms | 11.9 | 138 | 12,838 ms | 63% | 74% | 26.3x | 94/94 |
| llama4 · ollama · cold · 4 | 2 s | 64.0 GB | 3,035 ms | 11.2 | 136 | 13,714 ms | 63% | 76% | 28.1x | 94/94 |
| gemma4 · ollama · cold · 4 | 1 s | 4.0 GB | 757 ms | 43.3 | 485 | 10,037 ms | 62% | 70% | 20.5x | 94/94 |
| llama4 · 62.8 GB · ollama · cold · 1 | 17 s | 64.0 GB | 832 ms | 18.7 | 140 | 6,873 ms | 62% | 74% | 14.1x | 94/94 |
| llama4 · ollama · cached · 1 | 42 s | 64.0 GB | 553 ms | 18.7 | 139 | 6,784 ms | 62% | 74% | 13.9x | 94/94 |
| llama3:70b-instruct · 37.2 GB · ollama · cold · 1 | 12 s | 42.4 GB | 516 ms | 5.6 | 137 | 21,562 ms | 55% | 69% | 44.1x | 94/94 |
| llama3:70b-instruct · ollama · cached · 4 | 133 s | 42.4 GB | 1,137 ms | 5.3 | 134 | 24,921 ms | 55% | 68% | 51.0x | 94/94 |
| llama3:70b-instruct · ollama · cached · 1 | 136 s | 42.4 GB | 427 ms | 5.6 | 135 | 21,462 ms | 53% | 66% | 43.9x | 94/94 |
| llama3:70b-instruct · ollama · cold · 4 | 3 s | 42.4 GB | 937 ms | 5.3 | 136 | 26,618 ms | 53% | 67% | 54.5x | 94/94 |
| codellama:70b · ollama · cold · 4 | 3 s | 37.8 GB | 1,033 ms | 5.4 | 291 | 53,859 ms | 34% | 47% | 110.2x | 94/94 |
| codellama:70b · ollama · cached · 4 | 65 s | 37.8 GB | 1,207 ms | 5.4 | 290 | 52,649 ms | 34% | 45% | 107.7x | 94/94 |
| codellama:70b · 36.2 GB · ollama · cold · 1 | 11 s | 37.8 GB | 369 ms | 5.6 | 292 | 45,981 ms | 31% | 44% | 94.1x | 94/94 |
| codellama:70b · ollama · cached · 1 | 86 s | 37.8 GB | 266 ms | 5.6 | 291 | 46,229 ms | 30% | 42% | 94.6x | 94/94 |
| minicpm-v4.5 · 5.7 GB · ollama · cold · 1 | 8 s | 14.7 GB | 251 ms | 39.5 | 171 | 3,429 ms | 21% | 44% | 7.0x | 94/94 |
| minicpm-v4.5 · ollama · cold · 4 | 1 s | 14.7 GB | 428 ms | 34.4 | 168 | 4,081 ms | 20% | 41% | 8.4x | 94/94 |
| minicpm-v4.5 · ollama · cached · 1 | 15 s | 14.7 GB | 232 ms | 39.4 | 175 | 3,213 ms | 19% | 44% | 6.6x | 94/94 |
| minicpm-v4.5 · ollama · cached · 4 | 16 s | 14.7 GB | 417 ms | 33.9 | 179 | 4,443 ms | 19% | 42% | 9.1x | 94/94 |
| qwen3:4b · ollama · cold · 4 | 0 s | 12.6 GB | 558 ms | 59.8 | 1,003 | 17,690 ms | 13% | 11% | 36.2x | 94/94 |
| qwen3:4b · ollama · cached · 4 | 15 s | 12.6 GB | 542 ms | 59.9 | 985 | 17,647 ms | 12% | 10% | 36.1x | 94/94 |
| qwen3:4b · ollama · cached · 1 | 15 s | 12.6 GB | 262 ms | 71.5 | 1,000 | 14,582 ms | 11% | 9% | 29.8x | 94/94 |
| qwen3:4b · 2.3 GB · ollama · cold · 1 | 5 s | 12.6 GB | 247 ms | 71.5 | 1,000 | 14,576 ms | 11% | 9% | 29.8x | 94/94 |
| qwen3:0.6b · ollama · cached · 4 | 3 s | 8.5 GB | 328 ms | 207.8 | 138 | 760 ms | 4% | 30% | 1.6x | 94/94 |
| qwen3:0.6b · 0.5 GB · ollama · cold · 1 | 5 s | 8.5 GB | 204 ms | 295.0 | 127 | 519 ms | 3% | 31% | 1.1x | 94/94 |
| qwen3:0.6b · ollama · cold · 4 | 0 s | 8.5 GB | 332 ms | 210.6 | 140 | 706 ms | 3% | 29% | 1.4x | 94/94 |
| qwen3:0.6b · ollama · cached · 1 | 2 s | 8.5 GB | 198 ms | 294.1 | 130 | 489 ms | 2% | 30% | 1.0x | 94/94 |
Correctness alone is not a ranking — one point of correctness is not worth half the speed. Each model's best cell scores 70% × correctness + 30% × relative speed, where relative speed is decode rate against the fastest model in this report (qwen3:0.6b, 207.8 tok/s = 1.0). The bar is the weighting made visible: correctness · speed. Drag to change what you value; the ranking recomputes.
| # | Model | Correct | tok/s | Score |
|---|---|---|---|---|
| 1 | gemma4:26b | 87% | 63.2 | 70 |
| 2 | ornith-nvfp4 | 85% | 65.6 | 69 |
| 3 | qwen3-coder-next | 84% | 62.1 | 68 |
| 4 | qwen3.6:35b-a3b | 79% | 74.9 | 66 |
| 5 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS | 89% | 13.3 | 64 |
| 6 | gpt-oss:120b | 82% | 42.4 | 63 |
| 7 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M | 74% | 63.3 | 61 |
| 8 | qwen3-coder:30b | 70% | 83.7 | 61 |
| 9 | laguna-xs-2.1 | 74% | 52.7 | 60 |
| 10 | qwen3-coder:tuned | 72% | 60.7 | 59 |
| 11 | qwen3.6:27b | 79% | 12.1 | 57 |
| 12 | devstral-small-2:24b | 76% | 13.7 | 55 |
| 13 | devstral-2:123b | 77% | 2.6 | 54 |
| 14 | gemma3:27b | 72% | 11.5 | 52 |
| 15 | gemma4 | 65% | 42.4 | 52 |
| 16 | llama4 | 63% | 11.9 | 46 |
| 17 | llama3:70b-instruct | 55% | 5.6 | 40 |
| 18 | qwen3:0.6b | 4% | 207.8 | 33 |
| 19 | codellama:70b | 34% | 5.4 | 25 |
| 20 | minicpm-v4.5 | 21% | 39.5 | 21 |
| 21 | qwen3:4b | 13% | 59.8 | 18 |
Position is footprint against correctness; bubble area is output speed. Models whose size cannot be read — vLLM checkpoints live inside their containers — are absent, not zero.
The only controlled engine comparison a run can contain: identical weights reachable through more than one backend, paired by cache state. When output rates tie, the wait for the first token is what an engine buys.
Cold sends a uniquely salted prompt every time so nothing can be reused; cached repeats one identical prompt after a priming request. A backend whose prefix caching is off or unsupported shows roughly the same first-token latency in both columns — which looks like ordinary slowness rather than a misconfiguration.
| Model | Prompt | Think | Cold TTFT | Cached TTFT | Speed-up | |
|---|---|---|---|---|---|---|
DeepSeek-V4-Flash-0731-UD-IQ2_XXS | 0 | off | 1,461 ms | 1,426 ms | 1.0x | no measurable reuse |
gemma4:26b | 0 | off | 520 ms | 474 ms | 1.1x | no measurable reuse |
gemma4:26b | 0 | off | 829 ms | 833 ms | 1.0x | no measurable reuse |
DeepSeek-V4-Flash-0731-UD-IQ2_XXS | 0 | off | 340 ms | 338 ms | 1.0x | no measurable reuse |
ornith-nvfp4 | 0 | off | 118 ms | 117 ms | 1.0x | no measurable reuse |
qwen3-coder-next | 0 | off | 160 ms | 152 ms | 1.1x | no measurable reuse |
qwen3-coder-next | 0 | off | 405 ms | 424 ms | 1.0x | no measurable reuse |
ornith-nvfp4 | 0 | off | 325 ms | 324 ms | 1.0x | no measurable reuse |
gpt-oss:120b | 0 | off | 597 ms | 505 ms | 1.2x | no measurable reuse |
qwen3.6:35b-a3b | 0 | off | 4,687 ms | 4,847 ms | 1.0x | no measurable reuse |
qwen3.6:35b-a3b | 0 | off | 423 ms | 410 ms | 1.0x | no measurable reuse |
qwen3.6:27b | 0 | off | 608 ms | 619 ms | 1.0x | no measurable reuse |
qwen3.6:27b | 0 | off | 23,058 ms | 21,008 ms | 1.1x | no measurable reuse |
gpt-oss:120b | 0 | off | 1,327 ms | 1,277 ms | 1.0x | no measurable reuse |
devstral-2:123b | 0 | off | 740 ms | 692 ms | 1.1x | no measurable reuse |
devstral-2:123b | 0 | off | 2,696 ms | 2,885 ms | 0.9x | no measurable reuse |
devstral-small-2:24b | 0 | off | 348 ms | 340 ms | 1.0x | no measurable reuse |
Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M | 0 | off | 4,333 ms | 4,040 ms | 1.1x | no measurable reuse |
Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M | 0 | off | 406 ms | 399 ms | 1.0x | no measurable reuse |
laguna-xs-2.1:latest | 0 | off | 502 ms | 518 ms | 1.0x | no measurable reuse |
qwen3-coder:tuned | 0 | off | 4,216 ms | 4,080 ms | 1.0x | no measurable reuse |
qwen3-coder:tuned | 0 | off | 339 ms | 367 ms | 0.9x | no measurable reuse |
qwen3-coder-next:latest | 0 | off | 4,258 ms | 4,936 ms | 0.9x | no measurable reuse |
qwen3-coder-next:latest | 0 | off | 380 ms | 390 ms | 1.0x | no measurable reuse |
devstral-small-2:24b | 0 | off | 892 ms | 846 ms | 1.1x | no measurable reuse |
gemma3:27b | 0 | off | 655 ms | 632 ms | 1.0x | no measurable reuse |
gemma3:27b | 0 | off | 1,163 ms | 1,201 ms | 1.0x | no measurable reuse |
qwen3-coder:30b | 0 | off | 247 ms | 216 ms | 1.1x | no measurable reuse |
qwen3-coder:30b | 0 | off | 438 ms | 430 ms | 1.0x | no measurable reuse |
laguna-xs-2.1:latest | 0 | off | 261 ms | 184 ms | 1.4x | no measurable reuse |
gemma4:latest | 0 | off | 757 ms | 742 ms | 1.0x | no measurable reuse |
gemma4:latest | 0 | off | 479 ms | 470 ms | 1.0x | no measurable reuse |
llama4:latest | 0 | off | 3,035 ms | 2,943 ms | 1.0x | no measurable reuse |
llama4:latest | 0 | off | 832 ms | 553 ms | 1.5x | cache is working |
llama3:70b-instruct | 0 | off | 516 ms | 427 ms | 1.2x | no measurable reuse |
llama3:70b-instruct | 0 | off | 937 ms | 1,137 ms | 0.8x | no measurable reuse |
codellama:70b | 0 | off | 1,033 ms | 1,207 ms | 0.9x | no measurable reuse |
codellama:70b | 0 | off | 369 ms | 266 ms | 1.4x | no measurable reuse |
minicpm-v4.5:latest | 0 | off | 251 ms | 232 ms | 1.1x | no measurable reuse |
minicpm-v4.5:latest | 0 | off | 428 ms | 417 ms | 1.0x | no measurable reuse |
qwen3:4b | 0 | off | 558 ms | 542 ms | 1.0x | no measurable reuse |
qwen3:4b | 0 | off | 247 ms | 262 ms | 0.9x | no measurable reuse |
qwen3:0.6b | 0 | off | 332 ms | 328 ms | 1.0x | no measurable reuse |
qwen3:0.6b | 0 | off | 204 ms | 198 ms | 1.0x | no measurable reuse |
The warm-up sends the same prompt the measured runs use, so its first-token time is that prompt’s cold prefill; everything after it is served warm. A backend whose prefix caching is off or unsupported shows no gap between these two columns.
| Configuration | Prompt | Cold TTFT | Cached TTFT | Faster by |
|---|---|---|---|---|
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1 | 0 | 1,226 ms | 356 ms | 3x |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1 | 0 | 1,066 ms | 341 ms | 3x |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1 | 0 | 1,112 ms | 338 ms | 3x |
| llama4 · ollama · cached · 1 | 0 | 920 ms | 553 ms | 2x |
The core tier confirms a model is not broken; it saturates for anything capable, which is exactly why the hard tier exists. Compare two models on the hard row when both score 100% on core.
| Configuration | Core | Hard |
|---|---|---|
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4 | 100% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4 | 100% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4 | 100% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4 | 100% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4 | 100% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4 | 100% | 84% |
| gemma4:26b · ollama · cached · 1 | 87% | 88% |
| gemma4:26b · ollama · cached · 4 | 80% | 91% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1 | 93% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1 | 93% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1 | 93% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1 | 93% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1 | 93% | 84% |
| DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1 | 93% | 84% |
| gemma4:26b · 16.8 GB · ollama · cold · 1 | 83% | 88% |
| ornith-nvfp4 · NVFP4 · vllm · cold · 1 | 83% | 86% |
| gemma4:26b · ollama · cold · 4 | 80% | 88% |
| qwen3-coder-next · NVFP4 · vllm · cold · 1 | 93% | 80% |
| qwen3-coder-next · NVFP4 · vllm · cached · 1 | 90% | 81% |
| qwen3-coder-next · NVFP4 · vllm · cached · 4 | 90% | 81% |
| ornith-nvfp4 · NVFP4 · vllm · cached · 4 | 83% | 81% |
| gpt-oss:120b · 60.9 GB · ollama · cold · 1 | 83% | 81% |
| gpt-oss:120b · ollama · cached · 1 | 83% | 81% |
| ornith-nvfp4 · NVFP4 · vllm · cached · 1 | 77% | 83% |
| ornith-nvfp4 · NVFP4 · vllm · cold · 4 | 80% | 81% |
| qwen3-coder-next · NVFP4 · vllm · cold · 4 | 90% | 77% |
| qwen3.6:35b-a3b · ollama · cached · 4 | 87% | 75% |
| qwen3.6:35b-a3b · ollama · cold · 4 | 87% | 75% |
| qwen3.6:35b-a3b · ollama · cached · 1 | 87% | 75% |
| qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1 | 87% | 75% |
| qwen3.6:27b · 16.2 GB · ollama · cold · 1 | 73% | 81% |
| qwen3.6:27b · ollama · cached · 4 | 73% | 81% |
| qwen3.6:27b · ollama · cold · 4 | 73% | 81% |
| qwen3.6:27b · ollama · cached · 1 | 73% | 81% |
| gpt-oss:120b · ollama · cold · 4 | 87% | 72% |
| devstral-2:123b · 69.8 GB · ollama · cold · 1 | 87% | 72% |
| devstral-2:123b · ollama · cached · 4 | 87% | 72% |
| devstral-small-2:24b · 14.1 GB · ollama · cold · 1 | 80% | 73% |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4 | 73% | 75% |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1 | 73% | 75% |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4 | 73% | 75% |
| Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1 | 73% | 75% |
| laguna-xs-2.1 · ollama · cold · 4 | 70% | 77% |
| gpt-oss:120b · ollama · cached · 4 | 80% | 72% |
| devstral-2:123b · ollama · cached · 1 | 87% | 69% |
| devstral-2:123b · ollama · cold · 4 | 87% | 69% |
| laguna-xs-2.1 · ollama · cached · 4 | 73% | 73% |
| qwen3-coder:tuned · ollama · cold · 4 | 73% | 72% |
| qwen3-coder:tuned · 48.2 GB · ollama · cold · 1 | 73% | 72% |
| qwen3-coder:tuned · ollama · cached · 1 | 73% | 72% |
| qwen3-coder:tuned · ollama · cached · 4 | 73% | 72% |
| qwen3-coder-next · ollama · cached · 4 | 80% | 69% |
| qwen3-coder-next · 48.2 GB · ollama · cold · 1 | 80% | 69% |
| qwen3-coder-next · ollama · cold · 4 | 80% | 69% |
| qwen3-coder-next · ollama · cached · 1 | 80% | 69% |
| devstral-small-2:24b · ollama · cached · 4 | 80% | 69% |
| devstral-small-2:24b · ollama · cold · 4 | 77% | 70% |
| gemma3:27b · ollama · cached · 1 | 80% | 69% |
| gemma3:27b · 16.2 GB · ollama · cold · 1 | 80% | 69% |
| gemma3:27b · ollama · cold · 4 | 77% | 70% |
| devstral-small-2:24b · ollama · cached · 1 | 77% | 69% |
| gemma3:27b · ollama · cached · 4 | 77% | 69% |
| qwen3-coder:30b · 17.3 GB · ollama · cold · 1 | 67% | 72% |
| qwen3-coder:30b · ollama · cold · 4 | 77% | 66% |
| laguna-xs-2.1 · 18.9 GB · ollama · cold · 1 | 67% | 67% |
| qwen3-coder:30b · ollama · cached · 1 | 70% | 66% |
| qwen3-coder:30b · ollama · cached · 4 | 73% | 64% |
| laguna-xs-2.1 · ollama · cached · 1 | 67% | 66% |
| gemma4 · ollama · cached · 4 | 73% | 61% |
| gemma4 · ollama · cached · 1 | 63% | 64% |
| gemma4 · 8.9 GB · ollama · cold · 1 | 63% | 64% |
| llama4 · ollama · cached · 4 | 60% | 64% |
| llama4 · ollama · cold · 4 | 60% | 64% |
| gemma4 · ollama · cold · 4 | 63% | 61% |
| llama4 · 62.8 GB · ollama · cold · 1 | 60% | 62% |
| llama4 · ollama · cached · 1 | 60% | 62% |
| llama3:70b-instruct · 37.2 GB · ollama · cold · 1 | 50% | 58% |
| llama3:70b-instruct · ollama · cached · 4 | 50% | 58% |
| llama3:70b-instruct · ollama · cached · 1 | 47% | 56% |
| llama3:70b-instruct · ollama · cold · 4 | 50% | 55% |
| codellama:70b · ollama · cold · 4 | 33% | 34% |
| codellama:70b · ollama · cached · 4 | 37% | 33% |
| codellama:70b · 36.2 GB · ollama · cold · 1 | 30% | 31% |
| codellama:70b · ollama · cached · 1 | 30% | 30% |
| minicpm-v4.5 · 5.7 GB · ollama · cold · 1 | 20% | 22% |
| minicpm-v4.5 · ollama · cold · 4 | 20% | 20% |
| minicpm-v4.5 · ollama · cached · 1 | 20% | 19% |
| minicpm-v4.5 · ollama · cached · 4 | 17% | 20% |
| qwen3:4b · ollama · cold · 4 | 13% | 12% |
| qwen3:4b · ollama · cached · 4 | 13% | 11% |
| qwen3:4b · ollama · cached · 1 | 13% | 9% |
| qwen3:4b · 2.3 GB · ollama · cold · 1 | 13% | 9% |
| qwen3:0.6b · ollama · cached · 4 | 3% | 5% |
| qwen3:0.6b · 0.5 GB · ollama · cold · 1 | 0% | 5% |
| qwen3:0.6b · ollama · cold · 4 | 3% | 3% |
| qwen3:0.6b · ollama · cached · 1 | 0% | 3% |
Share of responses that passed every case for that task. A model strong everywhere except one task and a model mediocre throughout can share an overall average.
| Task | Perfect in | Configurations that missed it |
|---|---|---|
balanced_depthMax bracket nesting depth, -1 if unbalanced | 88 of 96 | qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
base_convertInteger between bases 2-36 with validation | 57 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · ollama · cached · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
bash_dedup_linesFirst occurrence wins, order preserved (sort -u reorders) | 83 of 96 | devstral-small-2:24b · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
bash_sort_versionsNumeric per-component version sort | 70 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
c_bit_count_rangePopcount over [a,b] — i <= INT_MAX loop overflow trap | 33 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
card_gridResponsive auto-fill card grid | 92 of 96 | qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4 |
clamp_addSaturating int addition — overflow trap | 76 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
clamp_mulSaturating int multiplication | 73 of 96 | laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
count_wordsCount words split on spaces and tabs | 83 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, minicpm-v4.5 · ollama · cached · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
cpp_wrap_countGreedy word-wrap line count — long words never split | 37 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
cs_round_halfHalf away from zero — Math.Round is banker's | 21 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
css_sticky_footerFlexbox sticky footer, no fixed positioning | 67 of 96 | gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
csv_escapeRFC 4180 CSV field quoting | 83 of 96 | minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
csv_lineSplit one CSV record honouring quotes | 27 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
data_tableRevenue table with caption and scoped headers | 88 of 96 | minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
dedent_textLongest common whitespace prefix — tabs are not spaces | 37 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4:26b · ollama · cold · 4, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
deep_getPath lookup like a.b[2].c, null on any miss | 19 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
expand_rangesExpand "1-3,7" specs with validation | 76 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · ollama · cold · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
glob_matchGlob matching with ? and * | 86 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
go_ipv4_validStrict dotted-quad IPv4 — leading zeros invalid | 16 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
go_rle_decodeDecode letter+count runs, strict validation | 50 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
group_rangesCollapse consecutive integers into range strings | 68 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
html_nav_currentNav list with aria-current on the active link | 88 of 96 | qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
interval_intersectIntersect two interval lists | 84 of 96 | minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
json_pointerResolve an RFC 6901 JSON Pointer | 76 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
justifyFull text justification | 62 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
login_formLogin form with labels bound to their inputs | 26 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4 |
lru_opsLRU cache with eviction order | 74 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
mid_floorFloor midpoint of two i64s — overflow and negatives | 19 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
ordinalEnglish ordinal suffix — the 11th/12th/13th trap | 31 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
parse_queryParse a URL query string into an object | 63 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · ollama · cached · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
path_normNormalise a POSIX path with . and .. | 76 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
php_flatten_keysNested assoc array to dot-notation keys | 68 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
pluckColumn from associative rows — null vs missing key | 72 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4 |
quoted_splitShell-style split honouring quotes and escapes | 49 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
ring_bufferRing buffer that overwrites the oldest on overflow | 79 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
roman_strictRoman to int, rejecting non-canonical forms | 4 of 96 | DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cached · 32,768 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 131,072 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · 84.6 GB · llamacpp · cold · 262,144 · 4, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 1, DeepSeek-V4-Flash-0731-UD-IQ2_XXS · IQ2_XXS · llamacpp · cold · 32,768 · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, gpt-oss:120b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder-next · NVFP4 · vllm · cold · 1, qwen3-coder-next · NVFP4 · vllm · cold · 4, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:27b · 16.2 GB · ollama · cold · 1, qwen3.6:27b · ollama · cached · 1, qwen3.6:27b · ollama · cached · 4, qwen3.6:27b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
round_toRound to nearest multiple, halves away from zero | 56 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gpt-oss:120b · ollama · cached · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
rust_kv_getk=v parsing: first '=' splits, last duplicate wins | 69 of 96 | devstral-2:123b · 69.8 GB · ollama · cold · 1, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cached · 4, devstral-2:123b · ollama · cold · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, llama4 · 62.8 GB · ollama · cold · 1, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
semver_cmpSemantic versions incl. pre-release precedence | 46 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cold · 4, gemma3:27b · 16.2 GB · ollama · cold · 1, gemma3:27b · ollama · cached · 1, gemma3:27b · ollama · cached · 4, gemma3:27b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, gemma4:26b · ollama · cached · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, qwen3-coder-next · NVFP4 · vllm · cached · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
slugifyURL slug: symbol runs become one hyphen | 64 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma4 · 8.9 GB · ollama · cold · 1, gemma4 · ollama · cached · 1, gemma4 · ollama · cached · 4, gemma4 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
snake_to_camelsnake_case to camelCase — digits stop capitalisation | 32 of 96 | Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · 20.6 GB · ollama · cold · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 1, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cached · 4, Qwen-AgentWorld-35B-A3B-GGUF:Q4_K_M · ollama · cold · 4, codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-small-2:24b · 14.1 GB · ollama · cold · 1, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, gemma4 · ollama · cold · 4, gemma4:26b · 16.8 GB · ollama · cold · 1, gemma4:26b · ollama · cached · 1, gemma4:26b · ollama · cached · 4, gemma4:26b · ollama · cold · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, gpt-oss:120b · ollama · cached · 4, laguna-xs-2.1 · 18.9 GB · ollama · cold · 1, laguna-xs-2.1 · ollama · cached · 1, laguna-xs-2.1 · ollama · cached · 4, laguna-xs-2.1 · ollama · cold · 4, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · 62.8 GB · ollama · cold · 1, llama4 · ollama · cached · 1, llama4 · ollama · cached · 4, llama4 · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, ornith-nvfp4 · NVFP4 · vllm · cached · 1, ornith-nvfp4 · NVFP4 · vllm · cached · 4, ornith-nvfp4 · NVFP4 · vllm · cold · 1, ornith-nvfp4 · NVFP4 · vllm · cold · 4, qwen3-coder-next · 48.2 GB · ollama · cold · 1, qwen3-coder-next · ollama · cached · 1, qwen3-coder-next · ollama · cached · 4, qwen3-coder-next · ollama · cold · 4, qwen3-coder:30b · 17.3 GB · ollama · cold · 1, qwen3-coder:30b · ollama · cached · 1, qwen3-coder:30b · ollama · cached · 4, qwen3-coder:30b · ollama · cold · 4, qwen3-coder:tuned · 48.2 GB · ollama · cold · 1, qwen3-coder:tuned · ollama · cached · 1, qwen3-coder:tuned · ollama · cached · 4, qwen3-coder:tuned · ollama · cold · 4, qwen3.6:35b-a3b · 22.3 GB · ollama · cold · 1, qwen3.6:35b-a3b · ollama · cached · 1, qwen3.6:35b-a3b · ollama · cached · 4, qwen3.6:35b-a3b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
sql_missing_usersAnti-join where NOT IN meets a NULL | 84 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
sql_top_spendersGROUP BY total with HAVING — not a WHERE | 90 of 96 | qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1 |
theme_varsDark-mode token inside a media query | 88 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4 |
tokenize_exprTokenise arithmetic, None on invalid input | 72 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, gemma4:26b · ollama · cached · 4, gpt-oss:120b · 60.9 GB · ollama · cold · 1, gpt-oss:120b · ollama · cached · 1, llama3:70b-instruct · 37.2 GB · ollama · cold · 1, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, llama4 · ollama · cached · 1, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
topo_lexSmallest topological order, None on cycle | 72 of 96 | codellama:70b · 36.2 GB · ollama · cold · 1, codellama:70b · ollama · cached · 1, codellama:70b · ollama · cached · 4, codellama:70b · ollama · cold · 4, devstral-2:123b · ollama · cached · 1, devstral-2:123b · ollama · cold · 4, devstral-small-2:24b · ollama · cached · 1, devstral-small-2:24b · ollama · cached · 4, devstral-small-2:24b · ollama · cold · 4, llama3:70b-instruct · ollama · cached · 1, llama3:70b-instruct · ollama · cached · 4, llama3:70b-instruct · ollama · cold · 4, minicpm-v4.5 · 5.7 GB · ollama · cold · 1, minicpm-v4.5 · ollama · cached · 1, minicpm-v4.5 · ollama · cached · 4, minicpm-v4.5 · ollama · cold · 4, qwen3:0.6b · 0.5 GB · ollama · cold · 1, qwen3:0.6b · ollama · cached · 1, qwen3:0.6b · ollama · cached · 4, qwen3:0.6b · ollama · cold · 4, qwen3:4b · 2.3 GB · ollama · cold · 1, qwen3:4b · ollama · cached · 1, qwen3:4b · ollama · cached · 4, qwen3:4b · ollama · cold · 4 |
The first failing case per configuration: the call that was made, what came back, and what should have — or the compile error or timeout that stopped it. This is what a percentage point of correctness is made of.
roman_strict — Roman to int, rejecting non-canonical forms · 4 examplesroman_strict("III") → null, expected 3def roman_strict(s: str) -> int | None:
# Define the values of Roman numeral symbols
values = {
'I': 1, 'V': 5, 'X': 10, 'L': 50,
'C': 100, 'D': 500, 'M': 1000
}
# Define subtractive pairs and their values
subtractive_pairs = {
'IV': 4, 'IX': 9, 'XL': 40, 'XC': 90,
'CD': 400, 'CM': 900
}
# Check for invalid characters
for ch in s:
if ch not in values:
return None
# Check for invalid repetitions
# I, X, C can repeat at most 3 times
for ch in 'IXC':
if s.count(ch) > 3:
return None
# V, L, D cannot repeat
for ch in 'VLD':
if s.count(ch) > 1:
return None
# Check for invalid subtractive pairs (each at most once)
for pair in subtractive_pairs:
if s.count(pair) > 1:
return None
# Check for invalid combinations like 'IIII', 'VV', 'IC', 'MCMC'
# Already covered by repetition rules, but also check for invalid subtractive patterns
# like 'IC', 'IM', 'XD', 'XM', etc.
invalid_subtractive = ['IC', 'IM', 'XD', 'XM', 'ID', 'II', 'VV', 'DD', 'LL']
for pattern in invalid_subtractive:
if pattern in s:
return None
# Check for non-increasing order (except subtractive pairs)
# We'll parse the string and check order
i = 0
n = len(s)
prev_value = float('inf')
used_subtractive = set()
while i < n:
# Check for subtractive pair
if i+1 < n and s[i:i+2] in subtractive_pairs:
pair = s[i:i+2]
… (truncated)roman_strict("MCMC") → 2000, expected nulldef roman_strict(s: str) -> int | None:
# Define the values of Roman numeral symbols
values = {
'I': 1, 'V': 5, 'X': 10, 'L': 50,
'C': 100, 'D': 500, 'M': 1000
}
# Define subtractive pairs and their values
subtractive_pairs = {
'IV': 4, 'IX': 9, 'XL': 40, 'XC': 90,
'CD': 400, 'CM': 900
}
# Check if all characters are valid
if not all(c in values for c in s):
return None
# Check for invalid repetitions
# I, X, C can repeat at most 3 times
for c in 'IXC':
if s.count(c) > 3:
return None
# V, L, D cannot repeat
for c in 'VLD':
if s.count(c) > 1:
return None
# Check subtractive pairs appear at most once
for pair in subtractive_pairs:
if s.count(pair) > 1:
return None
# Check that subtractive pairs are used correctly:
# They must be at the start of a "place" (i.e., not preceded by a smaller value)
# And they must not be followed by a larger value
# Also, symbols must be in non-increasing order except for subtractive pairs
# Parse the string to check order and compute value
i = 0
n = len(s)
total = 0
prev_value = float('inf') # start with a large value
while i < n:
# Check for subtractive pair
if i+1 < n and s[i:i+2] in subtractive_pairs:
pair = s[i:i+2]
pair_value = subtractive_pairs[pair]
# The pair must be valid: the first symbol must be less than the second
# (which is guarante
… (truncated)roman_strict("MCMC") → 2000, expected nulldef roman_strict(s: str) -> int | None:
# Define the values of Roman numeral symbols
values = {
'I': 1, 'V': 5, 'X': 10, 'L': 50,
'C': 100, 'D': 500, 'M': 1000
}
# Define subtractive pairs and their values
subtractive_pairs = {
'IV': 4, 'IX': 9, 'XL': 40, 'XC': 90,
'CD': 400, 'CM': 900
}
# Check if all characters are valid
if not all(c in values for c in s):
return None
# Check for invalid repetitions
# I, X, C can repeat at most 3 times
for c in 'IXC':
if s.count(c) > 3:
return None
# V, L, D cannot repeat
for c in 'VLD':
if s.count(c) > 1:
return None
# Check subtractive pairs appear at most once
for pair in subtractive_pairs:
if s.count(pair) > 1:
return None
# Check that subtractive pairs are used correctly:
# They must be at the start of a "place" (i.e., not preceded by a smaller value)
# And they must not be followed by a larger value
# Also, symbols must be in non-increasing order except for subtractive pairs
# Parse the string to check order and compute value
i = 0
n = len(s)
total = 0
prev_value = float('inf') # start with a large value
while i < n:
# Check for subtractive pair
if i+1 < n and s[i:i+2] in subtractive_pairs:
pair = s[i:i+2]
pair_value = subtractive_pairs[pair]
# The pair must be valid: the first symbol must be less than the second
# (which is guarante
… (truncated)roman_strict("MCMC") → 2000, expected nulldef roman_strict(s: str) -> int | None:
# Define the values of Roman numeral symbols
values = {
'I': 1, 'V': 5, 'X': 10, 'L': 50,
'C': 100, 'D': 500, 'M': 1000
}
# Define subtractive pairs and their values
subtractive_pairs = {
'IV': 4, 'IX': 9, 'XL': 40, 'XC': 90,
'CD': 400, 'CM': 900
}
# Check if all characters are valid
if not all(c in values for c in s):
return None
# Check for invalid repetitions
# I, X, C can repeat at most 3 times
for c in 'IXC':
if s.count(c) > 3:
return None
# V, L, D cannot repeat
for c in 'VLD':
if s.count(c) > 1:
return None
# Check subtractive pairs appear at most once
for pair in subtractive_pairs:
if s.count(pair) > 1:
return None
# Check that subtractive pairs are used correctly:
# They must be at the start of a "place" (i.e., not preceded by a smaller value)
# And they must not be followed by a larger value
# Also, symbols must be in non-increasing order except for subtractive pairs
# Parse the string to check order and compute value
i = 0
n = len(s)
total = 0
prev_value = float('inf') # start with a large value
while i < n:
# Check for subtractive pair
if i+1 < n and s[i:i+2] in subtractive_pairs:
pair = s[i:i+2]
pair_value = subtractive_pairs[pair]
# The pair must be valid: the first symbol must be less than the second
# (which is guarante
… (truncated)go_ipv4_valid — Strict dotted-quad IPv4 — leading zeros invalid · 4 examplescompile error: # command-line-arguments
./task.go:6:11: undefined: stringsfunc IPv4Valid(s string) bool {
parts := strings.Split(s, ".")
if len(parts) != 4 {
return false
}
for _, p := range parts {
if len(p) == 0 {
return false
}
if len(p) > 1 && p[0] == '0' {
return false
}
num := 0
for _, c := range p {
if c < '0' || c > '9' {
return false
}
num = num*10 + int(c-'0')
}
if num > 255 {
return false
}
}
return true
}
compile error: # command-line-arguments
./task.go:6:11: undefined: stringsfunc IPv4Valid(s string) bool {
parts := strings.Split(s, ".")
if len(parts) != 4 {
return false
}
for _, p := range parts {
if len(p) == 0 {
return false
}
if len(p) > 1 && p[0] == '0' {
return false
}
num := 0
for _, c := range p {
if c < '0' || c > '9' {
return false
}
num = num*10 + int(c-'0')
}
if num > 255 {
return false
}
}
return true
}
compile error: # command-line-arguments
./task.go:6:11: undefined: stringsfunc IPv4Valid(s string) bool {
parts := strings.Split(s, ".")
if len(parts) != 4 {
return false
}
for _, p := range parts {
if len(p) == 0 {
return false
}
if len(p) > 1 && p[0] == '0' {
return false
}
num := 0
for _, c := range p {
if c < '0' || c > '9' {
return false
}
num = num*10 + int(c-'0')
}
if num > 255 {
return false
}
}
return true
}
compile error: # command-line-arguments
./task.go:6:11: undefined: stringsfunc IPv4Valid(s string) bool {
parts := strings.Split(s, ".")
if len(parts) != 4 {
return false
}
for _, p := range parts {
if len(p) == 0 {
return false
}
if len(p) > 1 && p[0] == '0' {
return false
}
num := 0
for _, c := range p {
if c < '0' || c > '9' {
return false
}
num = num*10 + int(c-'0')
}
if num > 255 {
return false
}
}
return true
}
deep_get — Path lookup like a.b[2].c, null on any miss · 4 examplesdeepGet({"a": {"b": [{"c": 5}, {"c": 7}]}, "x": [1, [2, 3]], "n": null}, "x[1][0]") → null, expected 2function deepGet(obj, path) {
if (!obj || typeof path !== "string" || path.length === 0) return null;
const keys = path.split(".");
let current = obj;
for (let i = 0; i < keys.length; i++) {
const key = keys[i];
const bracketMatch = key.match(/^([^[]+)?\[(\d+)\]$/);
if (bracketMatch) {
const baseKey = bracketMatch[1];
const index = parseInt(bracketMatch[2], 10);
if (baseKey) {
if (current == null || typeof current[baseKey] === "undefined") return null;
current = current[baseKey];
}
if (current == null || !Array.isArray(current)) return null;
if (index < 0 || index >= current.length) return null;
current = current[index];
} else {
if (current == null || typeof current[key] === "undefined") return null;
current = current[key];
}
}
return current;
}
deepGet({"a": {"b": [{"c": 5}, {"c": 7}]}, "x": [1, [2, 3]], "n": null}, "x[1][0]") → null, expected 2function deepGet(obj, path) {
if (!obj || typeof path !== "string" || path.length === 0) return null;
const keys = path.split(".");
let current = obj;
for (let i = 0; i < keys.length; i++) {
const key = keys[i];
const bracketMatch = key.match(/^([^[]+)?\[(\d+)\]$/);
if (bracketMatch) {
const baseKey = bracketMatch[1];
const index = parseInt(bracketMatch[2], 10);
if (baseKey) {
if (current == null || typeof current[baseKey] === "undefined") return null;
current = current[baseKey];
}
if (current == null || !Array.isArray(current)) return null;
if (index < 0 || index >= current.length) return null;
current = current[index];
} else {
if (current == null || typeof current[key] === "undefined") return null;
current = current[key];
}
}
return current;
}
deepGet({"a": {"b": [{"c": 5}, {"c": 7}]}, "x": [1, [2, 3]], "n": null}, "x[1][0]") → null, expected 2function deepGet(obj, path) {
if (!obj || typeof path !== "string" || path.length === 0) return null;
const keys = path.split(".");
let current = obj;
for (let i = 0; i < keys.length; i++) {
const key = keys[i];
const bracketMatch = key.match(/^([^[]+)?\[(\d+)\]$/);
if (bracketMatch) {
const baseKey = bracketMatch[1];
const index = parseInt(bracketMatch[2], 10);
if (baseKey) {
if (current == null || typeof current[baseKey] === "undefined") return null;
current = current[baseKey];
}
if (current == null || !Array.isArray(current)) return null;
if (index < 0 || index >= current.length) return null;
current = current[index];
} else {
if (current == null || typeof current[key] === "undefined") return null;
current = current[key];
}
}
return current;
}
deepGet({"a": {"b": [{"c": 5}, {"c": 7}]}, "x": [1, [2, 3]], "n": null}, "x[1][0]") → null, expected 2function deepGet(obj, path) {
if (!obj || typeof path !== "string" || path.length === 0) return null;
const keys = path.split(".");
let current = obj;
for (let i = 0; i < keys.length; i++) {
const key = keys[i];
const bracketMatch = key.match(/^([^[]+)?\[(\d+)\]$/);
if (bracketMatch) {
const baseKey = bracketMatch[1];
const index = parseInt(bracketMatch[2], 10);
if (baseKey) {
if (current == null || typeof current[baseKey] === "undefined") return null;
current = current[baseKey];
}
if (current == null || !Array.isArray(current)) return null;
if (index < 0 || index >= current.length) return null;
current = current[index];
} else {
if (current == null || typeof current[key] === "undefined") return null;
current = current[key];
}
}
return current;
}
mid_floor — Floor midpoint of two i64s — overflow and negatives · 4 exampleshit the token cap mid-answer — mid_floor(-3, 0) → "0", expected -2fn mid_floor(a: i64, b: i64) -> i64 {
// To avoid overflow of (a + b), we use the property that
// floor((a + b) / 2) can be calculated using a/2 + b/2 and handling the remainder.
// However, a simpler way to handle both overflow and the floor logic is:
// If a and b have the same sign, (a + b) / 2 is safe if we use a/2 + b/2 + adjustment.
// But even better: Use the property that mid_floor(a, b) = floor(a/2 + b/2 + (a%2 + b%2)/2).
// To avoid all overflow and handle negative rounding:
if a >= 0 && b >= 0 {
// Standard case for positive numbers
a / 2 + b / 2 + (a % 2 + b % 2) / 2
} else if a < 0 && b < 0 {
// Standard case for negative numbers
// In Rust, -3 / 2 = -1. We need floor(-1.5) = -2.
let res = a / 2 + b / 2;
let rem_sum = (a % 2).abs() + (b % 2).abs();
// If there is a remainder and the sum is negative, we need to adjust for floor
if (a % 2 != 0 || b % 2 != 0) && (a + b < 0) {
// This part is tricky with overflow. Let's use a more robust approach:
// mid_floor(a, b) = floor((a+b)/2).
// If we can't add them, we use the identity:
// floor((a+b)/2) = a/2 + b/2 + (1 if a%2 != 0 and b%2 != 0 and signs are same else 0) ... no.
}
// Let's use the most robust method:
// mid_floor(a, b) is simply (a/2 + b/2) plus an adjustment if both are odd.
// If a and b are both odd, (a+b)/2 has a .5 remainder.
// If signs are same, floor(x.5) = x. If signs differ, floor(-x.5) = -x-1.
0 // place
… (truncated)hit the token cap mid-answer — compile error: error: this file contains an unclosed delimiter
--> /tmp/bench_rust_vp9xz3yq/task.rs:63:3
|
2 | fn mid_floor(a: i64, b: i64) -> i64 {
| fn mid_floor(a: i64, b: i64) -> i64 {
// To avoid overflow of (a + b), we use the property that
// floor((a + b) / 2) is equivalent to a/2 + b/2 + adjustment.
// However, a simpler way to handle all cases (including negatives and overflows)
// is to use the logic: if both have same sign, result is (a/2 + b/2).
// If different signs, we must be careful with truncation.
// A robust approach that avoids overflow:
// floor((a + b) / 2) == floor(a/2 + b/2 + (a%2 + b%2)/2)
// But even simpler: use the property of integer division in Rust.
// Rust's `/` truncates toward zero. We want floor (toward -inf).
// Let sum = a + b. We want floor(sum / 2).
// If sum is negative and odd, Rust's `sum / 2` returns -1 for -3/2, but we want -2.
// To avoid overflow of a + b, we can use the following:
// mid = (a / 2) + (b / 2) + remainder_adjustment
// where remainder_adjustment handles the bits lost by truncation.
let a_div = a / 2;
let b_div = b / 2;
let a_rem = a % 2; // -1, 0, or 1
let b_rem = b % 2; // -1, 0, or 1
// The true sum is (a_div * 2 + a_rem) + (b_div * 2 + b_rem)
// Which is 2 * (a_div + b_div) + (a_rem + b_rem)
// We want floor( [2 * (a_div + b_div) + (a_rem + b_rem)] / 2 )
// This simplifies to: (a_div + b_div) + floor((a_rem + b_rem) / 2.0)
let rem_sum = a_rem + b_rem;
let adjustment = if rem_sum < 0 {
// If rem_sum is -1 (e.g., -1 + 0), floor(-0.5) is -1
-1
} else if rem_sum == 0 {
0
} else {
// If rem_sum is 1
… (truncated)hit the token cap mid-answer — mid_floor(-3, 0) → "0", expected -2fn mid_floor(a: i64, b: i64) -> i64 {
// To avoid overflow of (a + b), we use the property that
// floor((a + b) / 2) can be calculated using a/2 + b/2 and handling the remainder.
// However, a simpler way to handle both overflow and the floor logic is:
// If a and b have the same sign, (a + b) / 2 is safe if we use a/2 + b/2 + adjustment.
// But even better: Use the property that mid_floor(a, b) = floor(a/2 + b/2 + (a%2 + b%2)/2).
// To avoid all overflow and handle negative rounding:
if a >= 0 && b >= 0 {
// Standard case for positive numbers
a / 2 + b / 2 + (a % 2 + b % 2) / 2
} else if a < 0 && b < 0 {
// Standard case for negative numbers
// In Rust, -3 / 2 = -1. We need floor(-1.5) = -2.
let res = a / 2 + b / 2;
let rem_sum = (a % 2).abs() + (b % 2).abs();
// If there is a remainder and the sum is negative, we need to adjust for floor
if (a % 2 != 0 || b % 2 != 0) && (a + b < 0) {
// This part is tricky with overflow. Let's use a more robust approach:
// mid_floor(a, b) = floor((a+b)/2).
// If we can't add them, we use the identity:
// floor((a+b)/2) = a/2 + b/2 + (1 if a%2 != 0 and b%2 != 0 and signs are same else 0) ... no.
}
// Let's use the most robust method:
// mid_floor(a, b) is simply (a/2 + b/2) plus an adjustment if both are odd.
// If a and b are both odd, (a+b)/2 has a .5 remainder.
// If signs are same, floor(x.5) = x. If signs differ, floor(-x.5) = -x-1.
0 // place
… (truncated)hit the token cap mid-answer — compile error: error: this file contains an unclosed delimiter
--> /tmp/bench_rust_8gzboqxj/task.rs:84:3
|
68 | fn mid_floor_final(a: i64, b: i64) -> i64 {
| fn mid_floor(a: i64, b: i64) -> i64 {
// To avoid overflow of (a + b), we use the property that
// floor((a + b) / 2) can be calculated using a/2 + b/2 and handling remainders.
// However, a simpler way to handle all signs and prevent overflow is:
// If a and b have the same sign, (a+b)/2 is safe if we use checked math or logic.
// A robust approach for any i64:
if a >= 0 && b >= 0 {
// Both positive: standard division works, but a + b might overflow.
// Use a/2 + b/2 + (a%2 + b%2)/2 logic or simply:
a / 2 + b / 2 + (a % 2 + b % 2) / 2
} else if a < 0 && b < 0 {
// Both negative: standard division truncates toward zero.
// We need to handle the floor logic for negative results.
let res = a / 2 + b / 2 + (a % 2 + b % 2) / 2;
// If there's a remainder that would push us towards negative infinity,
// but integer division truncated it toward zero, we adjust.
// Actually, the logic `a/2 + b/2 + (a%2 + b%2)/2` works for all signs
// if we ensure the "remainder" part handles the floor direction.
// Let's use a more universal approach:
let sum_rem = (a % 2).abs() + (b % 2).abs();
// This is getting complex. Let's use the simplest overflow-safe mid:
// floor((a+b)/2) == floor(a/2 + b/2 + (a%2 + b%2)/2)
// But we must handle the sign of the remainder carefully for floor.
// Correct approach:
// If a+b is even, result is (a+b)/2.
// If a+b is odd, result is (a+b-1)/2 if we want floor.
// To avoid
… (truncated)cs_round_half — Half away from zero — Math.Round is banker's · 4 examplescompile error: context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(6,35): error CS0103: The name 'MidpointRounding' does not exipublic static class Sol
{
public static int RoundHalf(double v)
{
// Math.Round with MidpointRounding.AwayFromZero rounds halves away from zero
return (int)Math.Round(v, MidpointRounding.AwayFromZero);
}
}
compile error: context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(6,35): error CS0103: The name 'MidpointRounding' does not exipublic static class Sol
{
public static int RoundHalf(double v)
{
// Math.Round with MidpointRounding.AwayFromZero rounds halves away from zero
return (int)Math.Round(v, MidpointRounding.AwayFromZero);
}
}
compile error: context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(6,35): error CS0103: The name 'MidpointRounding' does not exipublic static class Sol
{
public static int RoundHalf(double v)
{
// Math.Round with MidpointRounding.AwayFromZero rounds halves away from zero
return (int)Math.Round(v, MidpointRounding.AwayFromZero);
}
}
compile error: context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(6,35): error CS0103: The name 'MidpointRounding' does not exipublic static class Sol
{
public static int RoundHalf(double v)
{
// Math.Round with MidpointRounding.AwayFromZero rounds halves away from zero
return (int)Math.Round(v, MidpointRounding.AwayFromZero);
}
}
login_form — Login form with labels bound to their inputs · 4 examplesmethod of "form" → "POST", expected "post"<form action="/login" method="POST">
<div>
<label for="email">Email</label>
<input type="email" id="email" name="email" required>
</div>
<div>
<label for="password">Password</label>
<input type="password" id="password" name="password" required>
</div>
<button type="submit">Sign in</button>
</form>
method of "form" → "POST", expected "post"<form action="/login" method="POST">
<div>
<label for="email">Email</label>
<input type="email" id="email" name="email" required>
</div>
<div>
<label for="password">Password</label>
<input type="password" id="password" name="password" required>
</div>
<button type="submit">Sign in</button>
</form>
method of "form" → "POST", expected "post"<form method="POST" action="/login"> <label for="email">Email</label> <input type="email" id="email" name="email" required> <label for="password">Password</label> <input type="password" id="password" name="password" required> <button type="submit">Sign in</button> </form>
method of "form" → "POST", expected "post"<form method="POST" action="/login"> <label for="email">Email</label> <input type="email" id="email" name="email" required> <label for="password">Password</label> <input type="password" id="password" name="password" required> <button type="submit">Sign in</button> </form>
csv_line — Split one CSV record honouring quotes · 4 examplescsv_line("a,\"b,c\",d") → ["a", "bc,"], expected ["a", "b,c", "d"]def csv_line(line: str) -> list[str]:
if not line:
return ['']
fields = []
current_field = []
in_quotes = False
i = 0
n = len(line)
while i < n:
char = line[i]
if not in_quotes:
if char == '"' and (not current_field or (len(current_field) == 1 and current_field[0] == '')):
# Only a quote at the very start of a field triggers quoted mode
in_quotes = True
i += 1
continue
elif char == ',':
fields.append("".join(current_field))
current_field = []
i += 1
continue
else:
current_field.append(char)
else:
# Inside quotes
if i + 1 < n and char == '"' and line[i+1] == '"':
# Handle escaped double quotes ("")
current_field.append('"')
i += 2
continue
elif char == '"':
# Closing quote
in_quotes = False
i += 1
continue
else:
current_field.append(char)
i += 1
i += 1
fields.append("".join(current_field))
return fields
csv_line("a,\"b,c\",d") → ["a", "bc,"], expected ["a", "b,c", "d"]def csv_line(line: str) -> list[str]:
if not line:
return ['']
result = []
current_field = []
in_quotes = False
i = 0
n = len(line)
while i < n:
char = line[i]
if not in_quotes:
if char == '"' and (not current_field or (len(current_field) == 1 and current_field[0] == '')):
# Only a quote at the very start of a field triggers quoted mode
in_quotes = True
i += 1
continue
elif char == ',':
result.append("".join(current_field))
current_field = []
i += 1
continue
else:
current_field.append(char)
else:
# Inside quotes
if i + 1 < n and char == '"' and line[i+1] == '"':
# Handle escaped double quotes ("")
current_field.append('"')
i += 2
continue
elif char == '"':
# Closing quote
in_quotes = False
i += 1
continue
else:
current_field.append(char)
i += 1
i += 1
# Append the last field processed
result.append("".join(current_field))
return result
csv_line("a,b\"c,d") → ["a", "bc,d"], expected ["a", "b\"c", "d"]def csv_line(line: str) -> list[str]:
fields = []
current = ''
in_quotes = False
i = 0
while i < len(line):
c = line[i]
if in_quotes:
if c == '"':
if i + 1 < len(line) and line[i + 1] == '"':
current += '"'
i += 1
else:
in_quotes = False
else:
current += c
else:
if c == '"':
in_quotes = True
elif c == ',':
fields.append(current)
current = ''
else:
current += c
i += 1
fields.append(current)
return fields
csv_line("a,\"b,c\",d") → ["a", "bc,"], expected ["a", "b,c", "d"]def csv_line(line: str) -> list[str]:
if not line:
return ['']
fields = []
current_field = []
in_quotes = False
i = 0
n = len(line)
while i < n:
char = line[i]
if not in_quotes:
if char == '"' and (not current_field or (len(current_field) == 1 and current_field[0] == '')):
# Only a quote at the very start of a field triggers quoted mode
in_quotes = True
i += 1
continue
elif char == ',':
fields.append("".join(current_field))
current_field = []
i += 1
continue
else:
current_field.append(char)
else:
# Inside quotes
if i + 1 < n and char == '"' and line[i+1] == '"':
# Handle escaped double quotes ("")
current_field.append('"')
i += 2
continue
elif char == '"':
# Closing quote
in_quotes = False
i += 1
continue
else:
current_field.append(char)
i += 1
i += 1
# Append the last field processed
fields.append("".join(current_field))
return fields
ordinal — English ordinal suffix — the 11th/12th/13th trap · 4 examplescompile error: n/.cache/ai_proxy_cs/Program.cs(6,23): error CS0246: The type or namespace name 'ArgumentOutOfRangeException' could not be found (are you missing a using directive opublic static class Sol
{
public static string Ordinal(int n)
{
if (n < 0)
throw new ArgumentOutOfRangeException(nameof(n), "n must be non-negative");
// Special case for numbers ending in 11, 12, or 13
int lastTwoDigits = n % 100;
if (lastTwoDigits == 11 || lastTwoDigits == 12 || lastTwoDigits == 13)
return n + "th";
// Determine suffix based on last digit
int lastDigit = n % 10;
string suffix = lastDigit switch
{
1 => "st",
2 => "nd",
3 => "rd",
_ => "th"
};
return n + suffix;
}
}
compile error: n/.cache/ai_proxy_cs/Program.cs(6,23): error CS0246: The type or namespace name 'ArgumentOutOfRangeException' could not be found (are you missing a using directive opublic static class Sol
{
public static string Ordinal(int n)
{
if (n < 0)
throw new ArgumentOutOfRangeException(nameof(n), "n must be non-negative");
// Special case for numbers ending in 11, 12, or 13
int lastTwoDigits = n % 100;
if (lastTwoDigits == 11 || lastTwoDigits == 12 || lastTwoDigits == 13)
return n + "th";
// Determine suffix based on last digit
int lastDigit = n % 10;
string suffix = lastDigit switch
{
1 => "st",
2 => "nd",
3 => "rd",
_ => "th"
};
return n + suffix;
}
}
compile error: n/.cache/ai_proxy_cs/Program.cs(6,23): error CS0246: The type or namespace name 'ArgumentOutOfRangeException' could not be found (are you missing a using directive opublic static class Sol
{
public static string Ordinal(int n)
{
if (n < 0)
throw new ArgumentOutOfRangeException(nameof(n), "n must be non-negative");
// Special case for numbers ending in 11, 12, or 13
int lastTwoDigits = n % 100;
if (lastTwoDigits == 11 || lastTwoDigits == 12 || lastTwoDigits == 13)
return n + "th";
// Determine suffix based on last digit
int lastDigit = n % 10;
string suffix = lastDigit switch
{
1 => "st",
2 => "nd",
3 => "rd",
_ => "th"
};
return n + suffix;
}
}
compile error: n/.cache/ai_proxy_cs/Program.cs(6,23): error CS0246: The type or namespace name 'ArgumentOutOfRangeException' could not be found (are you missing a using directive opublic static class Sol
{
public static string Ordinal(int n)
{
if (n < 0)
throw new ArgumentOutOfRangeException(nameof(n), "n must be non-negative");
// Special case for numbers ending in 11, 12, or 13
int lastTwoDigits = n % 100;
if (lastTwoDigits == 11 || lastTwoDigits == 12 || lastTwoDigits == 13)
return n + "th";
// Determine suffix based on last digit
int lastDigit = n % 10;
string suffix = lastDigit switch
{
1 => "st",
2 => "nd",
3 => "rd",
_ => "th"
};
return n + suffix;
}
}
snake_to_camel — snake_case to camelCase — digits stop capitalisation · 4 examplessnake_to_camel("_leading") → "leading", expected "Leading"fn snake_to_camel(s: &str) -> String {
let mut result = String::with_capacity(s.len());
let mut capitalize_next = false;
let mut first_char_seen = false;
for c in s.chars() {
if c == '_' {
// If we have already processed characters, the next valid char should be uppercase
if first_char_seen {
capitalize_next = true;
}
} else {
if !first_char_seen {
// Handle leading underscores: skip them and don't capitalize the first real char
result.push(c);
first_char_seen = true;
} else if capitalize_next {
result.push(c.to_ascii_uppercase());
capitalize_next = false;
} else {
result.push(c);
}
}
}
result
}
snake_to_camel("_leading") → "leading", expected "Leading"fn snake_to_camel(s: &str) -> String {
let mut result = String::with_capacity(s.len());
let mut capitalize_next = false;
let mut first_char_found = false;
// We iterate through characters, but we need to handle the logic
// of skipping underscores and uppercasing the next valid char.
for c in s.chars() {
if c == '_' {
// If we encounter an underscore, the next non-underscore char should be uppercase
if first_char_found {
capitalize_next = true;
}
continue;
}
// If this is the very first non-underscore character,
// it stays lowercase (standard camelCase) or as is.
// The prompt implies leading underscores vanish and we only uppercase
// characters *after* an underscore.
if !first_char_found {
result.push(c);
first_char_found = true;
} else if capitalize_next {
result.extend(c.to_uppercase());
capitalize_next = false;
} else {
result.push(c);
}
}
result
}
snake_to_camel("_leading") → "leading", expected "Leading"fn snake_to_camel(s: &str) -> String {
let mut result = String::with_capacity(s.len());
let mut capitalize_next = false;
let mut first_char_found = false;
// We iterate through characters, but we need to handle the "leading underscores"
// by skipping them until we find a non-underscore character.
let chars: Vec<char> = s.chars().collect();
let mut i = 0;
// Skip leading underscores
while i < chars.len() && chars[i] == '_' {
i += 1;
}
// Determine the end index to handle trailing underscores
let mut last_valid_idx = chars.len();
while last_valid_idx > i && chars[last_valid_idx - 1] == '_' {
last_valid_idx -= 1;
}
for j in i..last_valid_idx {
let c = chars[j];
if c == '_' {
capitalize_next = true;
} else {
if capitalize_next {
result.push(c.to_ascii_uppercase());
capitalize_next = false;
} else {
result.push(c);
}
first_char_found = true;
}
}
result
}
snake_to_camel("_leading") → "leading", expected "Leading"fn snake_to_camel(s: &str) -> String {
let mut result = String::new();
let chars: Vec<char> = s.chars().collect();
let len = chars.len();
let mut i = 0;
// Skip leading underscores
while i < len && chars[i] == '_' {
i += 1;
}
// Process the rest
while i < len {
if chars[i] == '_' {
// Skip all consecutive underscores
while i < len && chars[i] == '_' {
i += 1;
}
// If there's a next character, uppercase it if it's alphabetic
if i < len {
let c = chars[i];
if c.is_ascii_alphabetic() {
result.push(c.to_ascii_uppercase());
} else {
result.push(c);
}
i += 1;
}
} else {
result.push(chars[i]);
i += 1;
}
}
// Remove trailing underscores (they would have been skipped already, but let's be safe)
// Actually, trailing underscores are already skipped at the start, and in the loop,
// if we have underscores at the end, they get skipped when we enter the while loop
// and then the condition `i < len` fails, so they're naturally ignored.
// But wait: what about underscores that are not followed by anything?
// They are just skipped. So no need to trim.
result
}
c_bit_count_range — Popcount over [a,b] — i <= INT_MAX loop overflow trap · 4 exampleshit the token cap mid-answer — compile error: argument 2 has type 'long long int' [-Wformat=]
84 | printf("%d\n", bit_count_range(2147483645, 2147483647));
| ~^ ~~~~~~~~~~~~~~~~~~~~long long bit_count_range(int a, int b) {
/**
* Helper function to count total set bits from 0 up to n (inclusive).
* Uses the property that at each bit position i (0-indexed),
* the pattern of 0s and 1s repeats every 2^(i+1) integers.
*/
auto long long count_to_n(long long n) {
if (n < 0) return 0;
long long total = 0;
long long range_size = n + 1;
for (int i = 0; i < 32; i++) {
long long period = 1LL << (i + 1);
long long full_periods = range_size / period;
total += full_periods * (period / 2);
long long remainder = range_size % period;
if (remainder > (period / 2)) {
total += (remainder - (period / 2));
}
}
return total;
}
// Total bits in [a, b] is count(b) - count(a-1)
// We use long long for calculations to prevent overflow during subtraction/logic
long long upper = (long long)b;
long long lower = (long long)a;
// Using a helper logic inside the function scope
// Since standard C doesn't support nested functions, we implement the logic inline.
auto count_bits_upto = [](long long n) -> long long {
long long count = 0;
for (int i = 0; i < 31; i++) {
long long bit_val = 1LL << i;
// Number of full blocks of size 2^(i+1)
long long full_blocks = (n + 1) / (bit_val * 2);
count += full_blocks * bit_val;
// Remaining bits in the incomplete block
long long remaining = (n + 1) % (bit
… (truncated)hit the token cap mid-answer — compile error: argument 2 has type 'long long int' [-Wformat=]
84 | printf("%d\n", bit_count_range(2147483645, 2147483647));
| ~^ ~~~~~~~~~~~~~~~~~~~~long long bit_count_range(int a, int b) {
/**
* Helper function to count total set bits from 0 up to n (inclusive).
* Uses the property that at each bit position i (0-indexed),
* the pattern of 0s and 1s repeats every 2^(i+1) integers.
*/
auto long long count_to_n(long long n) {
if (n < 0) return 0;
long long total = 0;
long long range_size = n + 1;
for (int i = 0; i < 32; i++) {
long long period = 1LL << (i + 1);
long long full_periods = range_size / period;
total += full_periods * (period / 2);
long long remainder = range_size % period;
if (remainder > (period / 2)) {
total += (remainder - (period / 2));
}
}
return total;
}
// Total bits in [a, b] is count(b) - count(a-1)
// We use long long for calculations to prevent overflow during subtraction/logic
long long upper = (long long)b;
long long lower = (long long)a;
// Using a helper logic inside the function scope
// Since standard C doesn't support nested functions, we implement the logic inline.
auto count_bits_upto = [](long long n) -> long long {
long long count = 0;
for (int i = 0; i < 31; i++) {
long long bit_val = 1LL << i;
// Number of full blocks of size 2^(i+1)
long long full_blocks = (n + 1) / (bit_val * 2);
count += full_blocks * bit_val;
// Remaining bits in the incomplete block
long long remaining = (n + 1) % (bit
… (truncated)timeout after 10.0slong long bit_count_range(int a, int b) {
// Count total set bits for all integers from 0 to n
// Then result is count(b) - count(a-1)
if (a > b) return 0;
// Function to count total set bits from 0 to n
long long count_set_bits_upto(int n) {
if (n < 0) return 0;
long long count = 0;
int i = 0;
while ((1 << i) <= n) {
// For each bit position i, count how many numbers from 0 to n have that bit set
int full_cycles = (n + 1) / (1 << (i + 1));
int remainder = (n + 1) % (1 << (i + 1));
int partial = remainder - (1 << i);
if (partial < 0) partial = 0;
count += (long long)full_cycles * (1 << i) + partial;
i++;
}
return count;
}
long long result = count_set_bits_upto(b) - count_set_bits_upto(a - 1);
return result;
}
bit_count_range(2147483645, 2147483647) → "-2147483557", expected 91long long bit_count_range(int a, int b) {
if (a > b) return 0;
// Helper function to count set bits from 0 to n
long long count_bits_to_n(int n) {
if (n <= 0) return 0;
long long count = 0;
long long i = 1;
while (i <= (long long)n) {
long long divisor = i * 2;
long long full_cycles = (long long)(n + 1) / divisor;
count += full_cycles * i;
long long remainder = (long long)(n + 1) % divisor;
if (remainder > i) {
count += remainder - i;
}
i <<= 1;
}
return count;
}
return count_bits_to_n(b) - count_bits_to_n(a - 1);
}
dedent_text — Longest common whitespace prefix — tabs are not spaces · 4 examplesdedent_text(" \n a\n b") → " \na\n b", expected "\na\n b"def dedent_text(s: str) -> str:
lines = s.split('\n')
# Find the longest common leading whitespace prefix
# Only consider non-empty, non-whitespace-only lines for the prefix
common_prefix = None
for line in lines:
if line and line != '':
# This line has content, consider its leading whitespace
leading = ''
for ch in line:
if ch in (' ', '\t'):
leading += ch
else:
break
if common_prefix is None:
common_prefix = leading
else:
# Find the common prefix between common_prefix and leading
# The common prefix must be a prefix of both
min_len = min(len(common_prefix), len(leading))
new_prefix = ''
for i in range(min_len):
if common_prefix[i] == leading[i]:
new_prefix += leading[i]
else:
break
common_prefix = new_prefix
# Now strip the common_prefix from each line
result = []
for line in lines:
if line and line != '':
# Strip the common_prefix from the beginning
if common_prefix and line.startswith(common_prefix):
result.append(line[len(common_prefix):])
else:
result.append(line)
else:
result.append('')
return '\n'.join(result)
dedent_text(" a\n\tb") → " a\nb", expected " a\n\tb"def dedent_text(s: str) -> str:
if not s:
return ""
lines = s.split('\n')
# Identify lines that are not empty or whitespace-only to find the common prefix
content_lines = [line for line in lines if line.strip()]
if not content_lines:
# If no content lines exist, all non-empty lines become empty strings
return '\n'.join(['' if not line.strip() else line for line in lines])
# Find the minimum common leading whitespace prefix among content lines
# We look at each character index of the shortest line
min_prefix_len = None
for line in content_lines:
# Count leading whitespace characters (spaces and tabs)
count = 0
for char in line:
if char in (' ', '\t'):
count += 1
else:
break
if min_prefix_len is None or count < min_prefix_len:
min_prefix_len = count
# If for some reason no prefix was found (shouldn't happen with content lines)
if min_prefix_len is None:
min_prefix_len = 0
result = []
for line in lines:
if not line.strip():
# Lines that are empty or whitespace-only become empty strings
result.append("")
else:
# Remove the common prefix from content lines
result.append(line[min_prefix_len:])
return '\n'.join(result)
dedent_text(" a\n\tb") → " a\nb", expected " a\n\tb"def dedent_text(s: str) -> str:
if not s:
return s
lines = s.split('\n')
# Find the minimum leading whitespace among non-empty lines
min_indent = float('inf')
for line in lines:
# Skip empty or whitespace-only lines
if not line.strip():
continue
# Count leading whitespace characters
indent = 0
for char in line:
if char in ' \t':
indent += 1
else:
break
if indent < min_indent:
min_indent = indent
# If no non-empty lines found, return original
if min_indent == float('inf'):
return s
# Remove the common leading whitespace from each line
result_lines = []
for line in lines:
# If line is empty or whitespace-only, make it empty
if not line.strip():
result_lines.append('')
continue
# Remove exactly min_indent characters from the beginning
result_lines.append(line[min_indent:])
return '\n'.join(result_lines)
dedent_text(" a\n\tb") → " a\nb", expected " a\n\tb"def dedent_text(s: str) -> str:
if not s:
return s
lines = s.split('\n')
# Find the minimum leading whitespace among non-empty lines
min_indent = float('inf')
for line in lines:
# Skip empty or whitespace-only lines for computing the minimum indent
if line.strip() == '':
continue
# Count leading whitespace characters (tabs and spaces are different)
indent = 0
for char in line:
if char in ' \t':
indent += 1
else:
break
if indent < min_indent:
min_indent = indent
# If no non-empty lines found, return original string
if min_indent == float('inf'):
return s
# Remove the common leading whitespace from all lines
result_lines = []
for line in lines:
# For empty or whitespace-only lines, make them empty
if line.strip() == '':
result_lines.append('')
else:
# Remove exactly min_indent characters from the beginning
result_lines.append(line[min_indent:])
return '\n'.join(result_lines)
cpp_wrap_count — Greedy word-wrap line count — long words never split · 4 exampleswrap_count("supercalifragilistic word", 10) → "1", expected 2#include <string>
#include <sstream>
int wrap_count(const std::string& s, int width) {
if (s.empty()) return 0;
std::istringstream iss(s);
std::string word;
int lines = 0;
int current_line_length = 0;
while (iss >> word) {
if (word.length() > width) {
// Word longer than width, placed alone on its own line
lines++;
current_line_length = 0;
} else {
if (lines == 0) {
lines = 1;
current_line_length = word.length();
} else if (current_line_length + 1 + word.length() <= width) {
// Word fits on current line
current_line_length += 1 + word.length();
} else {
// Word doesn't fit, start new line
lines++;
current_line_length = word.length();
}
}
}
return lines;
}
compile error: /tmp/bench_cpp_es1d67ym/task.cpp: In function 'int wrap_count(const std::string&, int)':
/tmp/bench_cpp_es1d67ym/task.cpp:77:29: error: variable 'std::istringstream int wrap_count(const std::string& s, int width) {
if (s.empty()) return 0;
int lines = 0;
size_t pos = 0;
size_t len = s.length();
while (pos < len) {
// Skip leading spaces
while (pos < len && s[pos] == ' ') {
pos++;
}
if (pos >= len) break;
// Find the next word
size_t word_start = pos;
while (pos < len && s[pos] != ' ') {
pos++;
}
size_t word_len = pos - word_start;
// Check if word fits on current line (or starts new line)
if (lines == 0 || word_len > width) {
// First word or word too long: starts new line
lines++;
// If word is longer than width, it still occupies one line
// and we move past it
} else {
// Check if word fits with previous words on current line
// We need to track current line length
// Since we don't track it explicitly, we'll use a different approach:
// Process words and simulate line building
}
}
// Better approach: simulate line building
lines = 0;
pos = 0;
len = s.length();
while (pos < len) {
// Skip leading spaces
while (pos < len && s[pos] == ' ') {
pos++;
}
if (pos >= len) break;
// Find next word
size_t word_start = pos;
while (pos < len && s[pos] != ' ') {
pos++;
}
size_t word_len = pos - word_start;
// Check if we need
… (truncated)compile error: /tmp/bench_cpp_5vkaxv07/task.cpp: In function 'int wrap_count(const std::string&, int)':
/tmp/bench_cpp_5vkaxv07/task.cpp:11:29: error: variable 'std::istringstream int wrap_count(const std::string& s, int width) {
if (s.empty()) return 0;
std::istringstream iss(s);
std::string word;
int lines = 0;
int current_line_length = 0;
while (iss >> word) {
if (word.length() > width) {
// Word longer than width goes on its own line
lines++;
current_line_length = 0;
} else if (current_line_length == 0) {
// First word on line
current_line_length = word.length();
} else if (current_line_length + 1 + word.length() <= width) {
// Word fits with a space
current_line_length += 1 + word.length();
} else {
// Start new line
lines++;
current_line_length = word.length();
}
}
// Count the last line if it has content
if (current_line_length > 0) {
lines++;
}
return lines;
}
compile error: ^
/tmp/bench_cpp_hcg2fwsb/task.cpp:17:81: error: expected primary-expression before '{' token
17 | #include <string>
#include <sstream>
int wrap_count(const std::string& s, int width)
{
if (s.empty()) return 0;
if (width <= 0) { // degenerate case: each word on its own line
std::istringstream ss(s);
return static_cast<int>(std::distance(std::istream_iterator<std::string>{ss},
std::istream_iterator<std::string>{}));
}
std::istringstream iss(s);
std::string word;
int lines = 0;
int cur_len = 0; // length of the line currently being built
while (iss >> word) {
int wlen = static_cast<int>(word.size());
if (wlen > width) { // word too long → its own line
if (cur_len != 0) ++lines; // finish previous line, if any
++lines; // the oversized word occupies a line alone
cur_len = 0;
continue;
}
if (cur_len == 0) { // start a new line with this word
cur_len = wlen;
} else {
// need one space before the next word
if (cur_len + 1 + wlen <= width) {
cur_len += 1 + wlen; // fits on current line
} else {
++lines; // close current line
cur_len = wlen; // start new line with this word
}
}
}
if (cur_len != 0) ++lines; // count the last partially‑filled line
return lines;
}
semver_cmp — Semantic versions incl. pre-release precedence · 4 examplessemver_cmp("1.0.0-alpha", "1.0.0") → 1, expected -1import re
def semver_cmp(a: str, b: str) -> int:
def parse_version(v: str):
# Remove build metadata (everything after '+')
base = v.split('+', 1)[0]
# Split version into core and pre-release parts
if '-' in base:
core_part, pre_release_part = base.split('-', 1)
pre_release = pre_release_part.split('.')
else:
core_part = base
pre_release = None
# Parse major, minor, patch
core = [int(x) for x in core_part.split('.')]
# Process pre-release identifiers
processed_pre = []
if pre_release is not None:
for part in pre_release:
if part.isdigit():
processed_pre.append((0, int(part))) # Numeric type 0
else:
processed_pre.append((1, part)) # String type 1
return core, processed_pre
core_a, pre_a = parse_version(a)
core_b, pre_b = parse_version(b)
# 1. Compare Major, Minor, Patch
for i in range(3):
if core_a[i] > core_b[i]: return 1
if core_a[i] < core_b[i]: return -1
# 2. Handle Pre-release precedence rule:
# A version WITH a pre-release tag has LOWER precedence than one WITHOUT.
if pre_a is not None and pre_b is None:
return -1
if pre_a is None and pre_b is not None:
return 1
if pre_a is None and pre_b is None:
return 0
# 3. Compare Pre-release identifiers
for part_a, part_b in zip(pre_a, pre_b):
# type_a/type_b (0 for int,
… (truncated)semver_cmp("1.0.0-alpha", "1.0.0") → 1, expected -1def semver_cmp(a: str, b: str) -> int:
def parse_version(version_str):
# Remove build metadata (after '+')
if '+' in version_str:
version_str = version_str.split('+')[0]
# Split into main version and pre-release
if '-' in version_str:
main_part, prerelease_part = version_str.split('-', 1)
else:
main_part = version_str
prerelease_part = None
# Parse main version (major.minor.patch)
parts = main_part.split('.')
major = int(parts[0])
minor = int(parts[1]) if len(parts) > 1 else 0
patch = int(parts[2]) if len(parts) > 2 else 0
# Parse pre-release identifiers
prerelease = []
if prerelease_part:
for identifier in prerelease_part.split('.'):
# Check if identifier is numeric
if identifier.isdigit():
prerelease.append((0, int(identifier)))
else:
prerelease.append((1, identifier))
return (major, minor, patch), prerelease
main_a, prerelease_a = parse_version(a)
main_b, prerelease_b = parse_version(b)
# Compare main version parts
for i in range(3):
if main_a[i] < main_b[i]:
return -1
elif main_a[i] > main_b[i]:
return 1
# If main versions are equal, compare pre-release info
# A version without pre-release has higher precedence than one with pre-release
if prerelease_a is None and prerelease_b is None:
… (truncated)semver_cmp("1.0.0-alpha", "1.0.0-alpha.1") → 0, expected -1def semver_cmp(a: str, b: str) -> int:
def parse_version(v: str):
# Split off build metadata
if '+' in v:
v = v.split('+', 1)[0]
# Split off pre-release
if '-' in v:
main_part, pre_part = v.split('-', 1)
pre = pre_part.split('.')
else:
main_part = v
pre = None
# Parse main version
parts = main_part.split('.')
major = int(parts[0])
minor = int(parts[1])
patch = int(parts[2])
return (major, minor, patch, pre)
def pre_release_key(pre):
if pre is None:
# No pre-release means higher precedence
return (1, [])
else:
# Has pre-release, lower precedence
# Each identifier: if all digits, compare numerically; else compare as string
identifiers = []
for ident in pre:
if ident.isdigit():
identifiers.append((0, int(ident), ''))
else:
identifiers.append((1, 0, ident))
return (0, identifiers)
ma = parse_version(a)
mb = parse_version(b)
# Compare major, minor, patch
for i in range(3):
if ma[i] < mb[i]:
return -1
elif ma[i] > mb[i]:
return 1
# Compare pre-release
pa = pre_release_key(ma[3])
pb = pre_release_key(mb[3])
# If one has pre-release and the other doesn't
if pa[0] == 1 and pb[0] == 0:
return 1 # a > b (no pre > pre)
if pa[
… (truncated)semver_cmp("1.0.0-beta.2", "1.0.0-beta.11") → -9, expected -1def semver_cmp(a: str, b: str) -> int:
def parse_version(version):
# Split into main version and pre-release/build parts
main_part = version.split('+')[0]
if '-' in main_part:
main, pre = main_part.split('-', 1)
build = None
else:
main, pre, build = main_part, None, None
# Parse major.minor.patch
major, minor, patch = map(int, main.split('.'))
return (major, minor, patch), pre, build
def compare_pre(pre_a, pre_b):
if not pre_a and not pre_b:
return 0
if not pre_a:
return 1
if not pre_b:
return -1
parts_a = pre_a.split('.')
parts_b = pre_b.split('.')
for i in range(min(len(parts_a), len(parts_b))):
part_a, part_b = parts_a[i], parts_b[i]
# Check if both are numeric
if part_a.isdigit() and part_b.isdigit():
cmp = int(part_a) - int(part_b)
if cmp != 0:
return cmp
else:
# Compare as strings (non-numeric ranks higher than numeric)
if part_a.isdigit():
return -1
elif part_b.isdigit():
return 1
else:
cmp = (part_a > part_b) - (part_a < part_b)
if cmp != 0:
return cmp
# If all compared parts are equal, longer pre-release is greater
return len(parts_a) - len(parts_b)
main_a, pre_a, _ = parse_version(a)
main_
… (truncated)quoted_split — Shell-style split honouring quotes and escapes · 4 exampleshit the token cap mid-answer — timeout after 10.0sdef quoted_split(s: str) -> list | None:
tokens = []
current_token = []
in_quotes = False
i = 0
n = len(s)
# Track if we are currently processing a token (to handle empty strings/whitespace)
# We need to know if the current sequence of characters is "active"
has_content = False
while i < n:
char = s[i]
if in_quotes:
if char == '\\':
if i + 1 < n and (s[i+1] == '"' or s[i+1] == '\\'):
current_token.append(s[i+1])
i += 2
has_content = True
continue
else:
# Treat backslash as literal if not escaping quote/backslash
current_token.append(char)
elif char == '"':
in_quotes = False
has_content = True
else:
current_token.append(char)
i += 1
else:
if char in (' ', '\t'):
if has_content or (not has_content and len(current_token) > 0):
# This handles the case where a token was just closed
tokens.append("".join(current_token))
current_token = []
has_content = False
elif i > 0 and s[i-1] == '"':
# Special case: "" is an empty token
# If we just finished a quote, it's already handled by the logic below
pass
i += 1
elif char == '"':
in_quotes = True
… (truncated)hit the token cap mid-answer — timeout after 10.0sdef quoted_split(s: str) -> list | None:
tokens = []
current_token = []
in_quotes = False
i = 0
n = len(s)
# Track if we are currently processing a token (to handle empty strings/whitespace)
# We need to know if the current sequence of characters is "active"
has_content = False
while i < n:
char = s[i]
if in_quotes:
if char == '\\':
if i + 1 < n and (s[i+1] == '"' or s[i+1] == '\\'):
current_token.append(s[i+1])
i += 2
has_content = True
continue
else:
# Treat backslash as literal if not escaping quote/backslash
current_token.append(char)
elif char == '"':
in_quotes = False
has_content = True
else:
current_token.append(char)
i += 1
else:
if char in (' ', '\t'):
if has_content or (not has_content and len(current_token) > 0):
# This handles the case where a token was just closed
tokens.append("".join(current_token))
current_token = []
has_content = False
elif i > 0 and s[i-1] == '"':
# Special case: "" is an empty token
# If we just finished a quote, it's already handled by the logic below
pass
i += 1
elif char == '"':
in_quotes = True
… (truncated)hit the token cap mid-answer — no code in responsehit the token cap mid-answer — no code in responsego_rle_decode — Decode letter+count runs, strict validation · 4 examplescompile error: # command-line-arguments
./task.go:6:16: undefined: strings
./task.go:14:13: undefined: isLetterfunc RleDecode(s string) string {
var result strings.Builder
i := 0
for i < len(s) {
// Expect a letter
if i >= len(s) {
return ""
}
c := s[i]
if !isLetter(c) {
return ""
}
i++
// Expect a non-empty count with no leading zeros
if i >= len(s) {
return ""
}
// Check for leading zero
if s[i] == '0' {
return ""
}
// Parse the count
count := 0
for i < len(s) && s[i] >= '0' && s[i] <= '9' {
d := int(s[i] - '0')
count = count*10 + d
i++
}
if count == 0 {
return ""
}
// Append the letter count times
for j := 0; j < count; j++ {
result.WriteByte(c)
}
}
return result.String()
}
hit the token cap mid-answer — no code in responsehit the token cap mid-answer — no code in responsecompile error: # command-line-arguments
./task.go:9:9: undefined: stringsfunc RleDecode(s string) string {
if len(s) == 0 {
return ""
}
var sb strings.Builder
i := 0
n := len(s)
for i < n {
// Expect a letter
ch := s[i]
if ch < 'a' || ch > 'z' {
return ""
}
i++
// Expect at least one digit
if i >= n {
return ""
}
// Check for leading zero
if s[i] == '0' {
return ""
}
// Parse the count
count := 0
for i < n && s[i] >= '0' && s[i] <= '9' {
digit := int(s[i] - '0')
count = count*10 + digit
i++
}
if count == 0 {
return ""
}
for j := 0; j < count; j++ {
sb.WriteByte(ch)
}
}
return sb.String()
}
round_to — Round to nearest multiple, halves away from zero · 4 examplesround_to(25, 10) → "20", expected 30int round_to(int n, int m) {
int q = n / m;
int r = n % m;
if (r * m < 0 && -r * m >= m / 2) {
return (q - (r < 0 ? 1 : 0)) * m;
}
return q * m;
}
round_to(-25, 10) → "-35", expected -30int round_to(int n, int m) {
if (m <= 0) return n;
// Handle positive and negative numbers separately to avoid truncation issues
if (n >= 0) {
return ((n + m / 2) / m) * m;
} else {
// For negative numbers, we need to round away from zero on exact halves
// n is negative, so let's work with absolute value conceptually
// We want: if n = -k*m - r where 0 <= r < m, then:
// if r == 0: result is n (already a multiple)
// if r > 0 and r < m/2: round toward zero -> n + r = n - (-r) ... let's think differently
// Alternative approach: use division that rounds toward negative infinity for the quotient
// For negative n, C division truncates toward zero, so n/m is like ceil(|n|/m) in magnitude but positive direction
// Let's compute the remainder properly
int abs_n = -n; // since n < 0
int r = abs_n % m;
if (r == 0) {
return n;
}
// Now we have n = -(q*m + r) where q >= 0, 0 < r < m
// We want to round away from zero on exact halves
// So if r > m/2 or (r == m/2 and m is even, i.e., exact half), we round further from zero
// If r < m/2, we round toward zero
// For exact half: r == m/2 requires m to be even
int threshold = m / 2;
if (r > threshold || (m % 2 == 0 && r == threshold)) {
// Round away from zero: make the magnitude larger
// n is negative, so we subtract more
return n - m;
} else
… (truncated)round_to(-25, 10) → "-35", expected -30int round_to(int n, int m) {
if (m <= 0) return n;
// Handle positive and negative numbers separately to avoid truncation issues
if (n >= 0) {
return ((n + m / 2) / m) * m;
} else {
// For negative numbers, we need to round away from zero on exact halves
// n is negative, so let's work with absolute value conceptually
// We want: if n = -k*m - r where 0 <= r < m, then:
// if r == 0: result is n (already a multiple)
// if r > 0 and r < m/2: round toward zero -> n + r = n - (-r) ... let's think differently
// Alternative approach: use division that rounds toward negative infinity for the quotient
// For negative n, C division truncates toward zero, so n/m is like ceil(|n|/m) in magnitude but positive direction
// Let's compute the remainder properly
int abs_n = -n; // since n < 0
int r = abs_n % m;
if (r == 0) {
return n;
}
// Now we have n = -(q*m + r) where q >= 0, 0 < r < m
// We want to round away from zero on exact halves
// So if r > m/2 or (r == m/2 and m is even, i.e., exact half), we round further from zero
// If r < m/2, we round toward zero
// For exact half: r == m/2 requires m to be even
int threshold = m / 2;
if (r > threshold || (m % 2 == 0 && r == threshold)) {
// Round away from zero: make the magnitude larger
// n is negative, so we subtract more
return n - m;
} else
… (truncated)round_to(-25, 10) → "-35", expected -30int round_to(int n, int m) {
if (m <= 0) return n;
// Handle positive and negative numbers separately to avoid truncation issues
if (n >= 0) {
return ((n + m / 2) / m) * m;
} else {
// For negative numbers, we need to round away from zero on exact halves
// n is negative, so let's work with absolute value conceptually
// We want: if n = -k*m - r where 0 <= r < m, then:
// if r == 0: result is n (already a multiple)
// if r > 0 and r < m/2: round toward zero -> n + r = n - (-r) ... let's think differently
// Alternative approach: use division that rounds toward negative infinity for the quotient
// For negative n, C division truncates toward zero, so n/m is like ceil(|n|/m) in magnitude but positive direction
// Let's compute the remainder properly
int abs_n = -n; // since n < 0
int r = abs_n % m;
if (r == 0) {
return n;
}
// Now we have n = -(q*m + r) where q >= 0, 0 < r < m
// We want to round away from zero on exact halves
// So if r > m/2 or (r == m/2 and m is even, i.e., exact half), we round further from zero
// If r < m/2, we round toward zero
// For exact half: r == m/2 requires m to be even
int threshold = m / 2;
if (r > threshold || (m % 2 == 0 && r == threshold)) {
// Round away from zero: make the magnitude larger
// n is negative, so we subtract more
return n - m;
} else
… (truncated)base_convert — Integer between bases 2-36 with validation · 4 exampleshit the token cap mid-answer — SyntaxError: invalid syntax (<model>, line 39)def base_convert(s: str, frm: int, to: int) -> str | None:
"""Convert an integer written in base `frm` to a string in base `to`.
Supports bases 2‑36 inclusive.
Input may have a leading '-', and any number of leading zeros.
Digits are 0‑9 then A‑Z (case‑insensitive). Output uses lowercase
letters, has no leading zeros, and zero is represented as "0"
without a sign.
Returns None for:
* empty input or input consisting only of a sign,
* characters not valid in the source base,
* bases outside 2..36.
"""
# ----- validate bases -------------------------------------------------
if not (2 <= frm <= 36) or not (2 <= to <= 36):
return None
if not s:
return None
# ----- sign handling --------------------------------------------------
negative = False
if s[0] == '-':
negative = True
s_body = s[1:]
else:
s_body = s
if not s_body: # only a sign, no digits
return None
# ----- digit validation & conversion to integer -----------------------
value = 0
for ch in s_body:
# map character to numeric value (case‑insensitive)
if '0' <= ch <= '9':
digit = ord(ch) - ord('0')
elifhit the token cap mid-answer — SyntaxError: invalid syntax (<model>, line 39)def base_convert(s: str, frm: int, to: int) -> str | None:
"""Convert an integer written in base `frm` to a string in base `to`.
Supports bases 2‑36 inclusive.
Input may have a leading '-', and any number of leading zeros.
Digits are 0‑9 then A‑Z (case‑insensitive). Output uses lowercase
letters, has no leading zeros, and zero is represented as "0"
without a sign.
Returns None for:
* empty input or input consisting only of a sign,
* characters not valid in the source base,
* bases outside 2..36.
"""
# ----- validate bases -------------------------------------------------
if not (2 <= frm <= 36) or not (2 <= to <= 36):
return None
if not s:
return None
# ----- sign handling --------------------------------------------------
negative = False
if s[0] == '-':
negative = True
s_body = s[1:]
else:
s_body = s
if not s_body: # only a sign, no digits
return None
# ----- digit validation & conversion to integer -----------------------
value = 0
for ch in s_body:
# map character to numeric value (case‑insensitive)
if '0' <= ch <= '9':
digit = ord(ch) - ord('0')
elifbase_convert("-1A", 16, 10) → "26", expected "-26"def base_convert(s: str, frm: int, to: int) -> str | None:
# Validate bases
if frm < 2 or frm > 36 or to < 2 or to > 36:
return None
if not s:
return None
# Check for sign
negative = False
start = 0
if s[0] == '-':
negative = True
start = 1
elif s[0] == '+':
start = 1
if start == len(s):
return None # Only sign
# Convert from base frm to integer
num = 0
for i in range(start, len(s)):
c = s[i]
if '0' <= c <= '9':
d = ord(c) - ord('0')
elif 'a' <= c.lower() <= 'z':
d = ord(c.lower()) - ord('a') + 10
else:
return None
if d >= frm:
return None
num = num * frm + d
if negative:
num = -num
# Convert integer to base to
if num == 0:
return '0'
digits = '0123456789abcdefghijklmnopqrstuvwxyz'
result = []
n = abs(num)
while n > 0:
result.append(digits[n % to])
n //= to
return ''.join(reversed(result))
hit the token cap mid-answer — SyntaxError: expected 'except' or 'finally' block (<model>, line 55)def base_convert(s: str, frm: int, to: int) -> str | None:
"""Convert an integer written in base `frm` to a string in base `to`.
Parameters
----------
s : str
Input representation (may start with '-', may contain leading zeros,
digits 0‑9 and letters a‑z/A‑Z).
frm, to : int
Bases, must be between 2 and 36 inclusive.
Returns
-------
str | None
The converted number using lowercase letters, no leading zeros,
sign only if the value is non‑zero and negative; ``None`` for any
invalid input.
"""
# ---- validate bases -------------------------------------------------
if not (2 <= frm <= 36) or not (2 <= to <= 36):
return None
# ---- empty string check ---------------------------------------------
if not s:
return None
# ---- sign handling --------------------------------------------------
negative = False
if s[0] == '-':
negative = True
s_body = s[1:]
else:
s_body = s
# after removing a possible sign the body must contain at least one char
if not s_body:
return None
# ---- character validation & conversion to integer -------------------
def char_value(ch: str) -> int | None:
"""Return numeric value of a digit/letter, or None if invalid."""
if '0' <= ch <= '9':
return ord(ch) - 48
lo = ch.lower()
if 'a' <= lo <= 'z':
return ord(lo) - 87 # 'a' -> 10
return None
for c in s_body:
val = char_value(c)
if val is None or
… (truncated)justify — Full text justification · 4 examplesjustify(["aa", "bb"], 5) → ["aa ", "bb "], expected ["aa bb"]def justify(words: list, width: int) -> list:
result = []
i = 0
n = len(words)
while i < n:
# Find how many words can fit in current line
line_words = []
line_length = 0
while i < n and line_length + len(words[i]) + (1 if line_words else 0) <= width:
line_words.append(words[i])
line_length += len(words[i]) + (1 if line_words else 0)
i += 1
# Build the line
if i == n or len(line_words) == 1:
# Last line or single word line: left-justified
line = ' '.join(line_words)
line += ' ' * (width - len(line))
else:
# Distribute spaces evenly with extra spaces on left gaps
total_spaces = width - sum(len(word) for word in line_words)
gaps = len(line_words) - 1
base_spaces = total_spaces // gaps
extra_spaces = total_spaces % gaps
# Build line with extra spaces on left gaps
line_parts = []
for j, word in enumerate(line_words):
line_parts.append(word)
if j < len(line_words) - 1:
# Add base spaces + extra spaces for left gaps
spaces = base_spaces + (1 if j < extra_spaces else 0)
line_parts.append(' ' * spaces)
line = ''.join(line_parts)
result.append(line)
return result
justify(["This", "is", "an", "example", "of", "text", "justification."], 16) → [" This is an", " example of text", "justification. "], expected ["This is an", "example of text", "justification. "]def justify(words: list, width: int) -> list:
"""
Perform full text justification.
Pack as many words per line as fit with at least one space between them.
Pad each line to exactly width by distributing spaces between words as evenly as possible,
putting the extra spaces on the LEFT gaps first.
The last line, and any line holding a single word, is left justified with the remaining space on the right.
"""
if not words:
return []
# Group words into lines
lines = []
current_line = []
current_length = 0
for word in words:
# Check if adding this word exceeds the width
# At least one space between words, so new length = current_length + 1 (space) + len(word)
if current_line and current_length + 1 + len(word) > width:
lines.append(current_line)
current_line = [word]
current_length = len(word)
else:
current_line.append(word)
current_length += (1 + len(word)) if current_line else len(word)
if current_line:
lines.append(current_line)
# Now justify each line except the last
result = []
for i, line in enumerate(lines):
if i == len(lines) - 1:
# Last line: left-justified
result.append(' '.join(line).ljust(width))
elif len(line) == 1:
# Single word: left-justified
result.append(line[0].ljust(width))
else:
# Full justification
total_words_len = sum(len(w) for w in line)
gaps = len(line
… (truncated)hit the token cap mid-answer — SyntaxError: '(' was never closed (<model>, line 46)def justify(words: list, width: int) -> list:
"""
Full‑justify a list of words into lines of exactly *width* characters.
Rules
-----
1. Pack as many words per line as fit, leaving at least one space between them.
2. For every line except the last **and** lines that contain only one word:
distribute the required spaces as evenly as possible; any leftover spaces
are placed in the leftmost gaps.
3. The last line and any single‑word line are left‑justified: words are joined
by a single space and the remaining blanks are appended on the right.
Parameters
----------
words : list[str]
List of words to be justified (already stripped of whitespace).
width : int
Desired line length; all returned strings have exactly this length.
Returns
-------
list[str]
The justified lines.
"""
res = []
i, n = 0, len(words)
while i < n:
# ---- determine the words that fit on the current line ----
line_len = len(words[i]) # length of first word
j = i + 1 # index after the last word in this line
while j < n and line_len + 1 + len(words[j]) <= width:
line_len += 1 + len(words[j]) # add one space plus next word
j += 1
line_words = words[i:j]
is_last_line = (j == n)
# ---- build the line ------------------------------------
if len(line_words) == 1 or is_last_line:
# left‑justified: single spaces between words, pad on the right
line = " ".j
… (truncated)hit the token cap mid-answer — SyntaxError: '(' was never closed (<model>, line 46)def justify(words: list, width: int) -> list:
"""
Full‑justify a list of words into lines of exactly *width* characters.
Rules
-----
1. Pack as many words per line as fit, leaving at least one space between them.
2. For every line except the last **and** lines that contain only one word:
distribute the required spaces as evenly as possible; any leftover spaces
are placed in the leftmost gaps.
3. The last line and any single‑word line are left‑justified: words are joined
by a single space and the remaining blanks are appended on the right.
Parameters
----------
words : list[str]
List of words to be justified (already stripped of whitespace).
width : int
Desired line length; all returned strings have exactly this length.
Returns
-------
list[str]
The justified lines.
"""
res = []
i, n = 0, len(words)
while i < n:
# ---- determine the words that fit on the current line ----
line_len = len(words[i]) # length of first word
j = i + 1 # index after the last word in this line
while j < n and line_len + 1 + len(words[j]) <= width:
line_len += 1 + len(words[j]) # add one space plus next word
j += 1
line_words = words[i:j]
is_last_line = (j == n)
# ---- build the line ------------------------------------
if len(line_words) == 1 or is_last_line:
# left‑justified: single spaces between words, pad on the right
line = " ".j
… (truncated)parse_query — Parse a URL query string into an object · 4 examplesparseQuery("a=%3D=&b") → {"a": "=", "b": ""}, expected {"a": "==", "b": ""}function parseQuery(qs) {
const result = {};
if (!qs || qs.length === 0) return result;
// Remove leading '?' if present
const queryString = qs.startsWith('?') ? qs.slice(1) : qs;
// Split into pairs
const pairs = queryString.split('&');
for (const pair of pairs) {
if (!pair) continue; // Skip empty pairs
// Split on first '=' only
const [key, value] = pair.split('=', 2);
// Decode the key and value
const decode = (str) => str.replace(/\+/g, ' ').replace(/%[0-9a-f]{2}/gi, char =>
String.fromCharCode(parseInt(char.slice(1), 16))
);
const decodedKey = decode(key);
const decodedValue = value !== undefined ? decode(value) : '';
// Handle repeated keys
if (result.hasOwnProperty(decodedKey)) {
const existing = result[decodedKey];
if (Array.isArray(existing)) {
existing.push(decodedValue);
} else {
result[decodedKey] = [existing, decodedValue];
}
} else {
result[decodedKey] = decodedValue;
}
}
return result;
}
ReferenceError: parseQuery is not defined/**
* Parses a URL query string into a plain object.
*
* @param {string} qs The query string to parse.
* @returns {object} A plain object representing the parsed query string.
*/
const parseQuery = (qs) => {
if (!qs) {
return {};
}
if (qs.startsWith('?')) {
qs = qs.slice(1);
}
const result = {};
const pairs = qs.split('&');
for (const pair of pairs) {
if (!pair) {
continue;
}
const [key, value] = pair.split('=', 1);
let decodedKey = key ? decodeURIComponent(key.replace(/\+/g, ' ')) : '';
let decodedValue = value !== undefined ? decodeURIComponent(value.replace(/\+/g, ' ')) : '';
if (result.hasOwnProperty(decodedKey)) {
if (!Array.isArray(result[decodedKey])) {
result[decodedKey] = [result[decodedKey]];
}
result[decodedKey].push(decodedValue);
} else {
result[decodedKey] = decodedValue;
}
}
return result;
};
ReferenceError: parseQuery is not defined/**
* Parses a URL query string into a plain object.
*
* @param {string} qs The query string to parse.
* @returns {object} A plain object representing the parsed query string.
*/
const parseQuery = (qs) => {
if (!qs) {
return {};
}
if (qs.startsWith('?')) {
qs = qs.slice(1);
}
const result = {};
const pairs = qs.split('&');
for (const pair of pairs) {
if (!pair) {
continue;
}
const [key, value] = pair.split('=', 1);
let decodedKey = key ? decodeURIComponent(key.replace(/\+/g, ' ')) : '';
let decodedValue = value !== undefined ? decodeURIComponent(value.replace(/\+/g, ' ')) : '';
if (result.hasOwnProperty(decodedKey)) {
if (!Array.isArray(result[decodedKey])) {
result[decodedKey] = [result[decodedKey]];
}
result[decodedKey].push(decodedValue);
} else {
result[decodedKey] = decodedValue;
}
}
return result;
};
ReferenceError: parseQuery is not defined/**
* Parses a URL query string into a plain object.
*
* @param {string} qs The query string to parse.
* @returns {object} A plain object representing the parsed query string.
*/
const parseQuery = (qs) => {
if (!qs) {
return {};
}
if (qs.startsWith('?')) {
qs = qs.slice(1);
}
const result = {};
const pairs = qs.split('&');
for (const pair of pairs) {
if (!pair) {
continue;
}
const [key, value] = pair.split('=', 1);
let decodedKey = key ? decodeURIComponent(key.replace(/\+/g, ' ')) : '';
let decodedValue = value !== undefined ? decodeURIComponent(value.replace(/\+/g, ' ')) : '';
if (result.hasOwnProperty(decodedKey)) {
if (!Array.isArray(result[decodedKey])) {
result[decodedKey] = [result[decodedKey]];
}
result[decodedKey].push(decodedValue);
} else {
result[decodedKey] = decodedValue;
}
}
return result;
};
slugify — URL slug: symbol runs become one hyphen · 4 examplesslugify("Hello, World!") → "$__cases = json_decode(base64_decode(\"W1siSGVsbG8sIFdvcmxkIS\u2026", expected "hello-world"<?php
function slugify(string $s): string {
$s = strtolower($s);
$s = preg_replace('/[^a-z0-9]+/', '-', $s);
$s = trim($s, '-');
return $s;
}
?>
slugify("Hello, World!") → "$__cases = json_decode(base64_decode(\"W1siSGVsbG8sIFdvcmxkIS\u2026", expected "hello-world"<?php
function slugify(string $s): string {
$s = strtolower($s);
$s = preg_replace('/[^a-z0-9]+/', '-', $s);
$s = trim($s, '-');
return $s;
}
?>
slugify("Hello, World!") → "$__cases = json_decode(base64_decode(\"W1siSGVsbG8sIFdvcmxkIS\u2026", expected "hello-world"<?php
function slugify(string $s): string {
$s = strtolower($s);
$s = preg_replace('/[^a-z0-9]+/', '-', $s);
$s = trim($s, '-');
return $s;
}
?>
slugify("Hello, World!") → "$__cases = json_decode(base64_decode(\"W1siSGVsbG8sIFdvcmxkIS\u2026", expected "hello-world"<?php
function slugify(string $s): string {
$s = strtolower($s);
$s = preg_replace('/[^a-z0-9]+/', '-', $s);
$s = trim($s, '-');
return $s;
}
?>
css_sticky_footer — Flexbox sticky footer, no fixed positioning · 4 examplesfooter { margin-top } → null, expected "auto"body {
display: flex;
flex-direction: column;
min-height: 100vh;
margin: 0;
}
main {
flex: 1;
}
footer {
/* footer naturally stays at the bottom due to flexbox layout */
}
footer { margin-top } → null, expected "auto"html, body {
height: 100%;
margin: 0;
}
body {
display: flex;
flex-direction: column;
min-height: 100vh;
}
main {
flex: 1;
}
footer {
/* Footer stays at the bottom because main grows to fill available space */
}
footer { margin-top } → null, expected "auto"html, body {
height: 100%;
margin: 0;
}
body {
display: flex;
flex-direction: column;
min-height: 100vh;
}
main {
flex: 1;
}
footer {
/* Footer stays at the bottom because main grows to fill available space */
}
footer { margin-top } → null, expected "auto"html, body {
height: 100%;
margin: 0;
}
body {
display: flex;
flex-direction: column;
min-height: 100vh;
}
main {
flex: 1;
}
footer {
/* Footer stays at the bottom because main grows to fill available space */
}
php_flatten_keys — Nested assoc array to dot-notation keys · 4 examplesflatten_keys({"x": {"y": {"z": "deep"}}}) → "", expected {"x.y.z": "deep"}function flatten_keys(array $a): array
{
$result = [];
function _flatten(array $arr, string $prefix, array &$result): void
{
foreach ($arr as $key => $value) {
$newKey = $prefix === '' ? (string)$key : $prefix . '.' . $key;
if (is_array($value)) {
_flatten($value, $newKey, $result);
} else {
$result[$newKey] = $value;
}
}
}
_flatten($a, '', $result);
return $result;
}
flatten_keys({"x": {"y": {"z": "deep"}}}) → "", expected {"x.y.z": "deep"}function flatten_keys(array $a): array {
$result = [];
function _flatten($data, string $prefix, array &$result): void {
foreach ($data as $key => $value) {
$newKey = $prefix === '' ? (string)$key : $prefix . '.' . $key;
if (is_array($value)) {
_flatten($value, $newKey, $result);
} else {
$result[$newKey] = $value;
}
}
}
_flatten($a, '', $result);
return $result;
}
flatten_keys({"a": {"b": 1, "c": {"d": 2}}, "e": 3}) → "$__cases = json_decode(base64_decode(\"W1t7ImEiOiB7ImIiOiAxLC\u2026", expected {"a.b": 1, "a.c.d": 2, "e": 3}<?php
function flatten_keys(array $a): array {
$recurse = function (array $arr, string $prefix) use (&$recurse): array {
$result = [];
foreach ($arr as $key => $value) {
$fullKey = $prefix === '' ? (string)$key : $prefix . '.' . $key;
if (is_array($value)) {
$result += $recurse($value, $fullKey);
} else {
$result[$fullKey] = $value;
}
}
return $result;
};
return $recurse($a, '');
}
?>
flatten_keys({"a": {"b": 1, "c": {"d": 2}}, "e": 3}) → "$__cases = json_decode(base64_decode(\"W1t7ImEiOiB7ImIiOiAxLC\u2026", expected {"a.b": 1, "a.c.d": 2, "e": 3}<?php
function flatten_keys(array $a): array {
$result = [];
$stack = [[$a, '']];
while ($stack) {
[$current, $prefix] = array_pop($stack);
foreach ($current as $key => $value) {
$newKey = $prefix === '' ? (string)$key : $prefix . '.' . (string)$key;
if (is_array($value)) {
$stack[] = [$value, $newKey];
} else {
$result[$newKey] = $value;
}
}
}
return $result;
}
?>
group_ranges — Collapse consecutive integers into range strings · 4 examplesgroupRanges([1, 2, 3, 7, 9, 10, 11]) → "1,3-2,7-3,9-7,10-9,11-10", expected "1-3,7,9-11"function groupRanges(nums) {
if (nums.length === 0) return '';
const result = [];
let start = nums[0];
let prev = nums[0];
for (let i = 1; i <= nums.length; i++) {
// When we reach the end or find a non-consecutive number
if (i === nums.length || nums[i] !== prev + 1) {
if (start === prev) {
result.push(String(start));
} else {
result.push(`${start}-${prev}`);
}
// Reset start for the next range
if (i < nums.length) {
start = nums[i];
}
}
prev = nums[i - 1];
}
return result.join(',');
}
groupRanges([1, 2, 3, 7, 9, 10, 11]) → "1,3-2,7-3,9-7,10-9,11-10", expected "1-3,7,9-11"function groupRanges(nums) {
if (nums.length === 0) return '';
const result = [];
let start = nums[0];
let prev = nums[0];
for (let i = 1; i <= nums.length; i++) {
// When we reach the end or find a non-consecutive number
if (i === nums.length || nums[i] !== prev + 1) {
if (start === prev) {
result.push(String(start));
} else {
result.push(`${start}-${prev}`);
}
// Reset start for the next range
if (i < nums.length) {
start = nums[i];
}
}
prev = nums[i - 1];
}
return result.join(',');
}
groupRanges([1, 2, 3, 7, 9, 10, 11]) → "1,3-2,7-3,9-7,10-9,11-10", expected "1-3,7,9-11"function groupRanges(nums) {
if (nums.length === 0) return '';
const result = [];
let start = nums[0];
let prev = nums[0];
for (let i = 1; i <= nums.length; i++) {
// When we reach the end or find a non-consecutive number
if (i === nums.length || nums[i] !== prev + 1) {
if (start === prev) {
result.push(String(start));
} else {
result.push(`${start}-${prev}`);
}
// Reset start for the next range
if (i < nums.length) {
start = nums[i];
}
}
prev = nums[i - 1];
}
return result.join(',');
}
groupRanges([1, 2, 3, 7, 9, 10, 11]) → "1,3-2,7-3,9-7,10-9,11-10", expected "1-3,7,9-11"function groupRanges(nums) {
if (nums.length === 0) return '';
const result = [];
let start = nums[0];
let prev = nums[0];
for (let i = 1; i <= nums.length; i++) {
// When we reach the end or find a non-consecutive number
if (i === nums.length || nums[i] !== prev + 1) {
if (start === prev) {
result.push(String(start));
} else {
result.push(`${start}-${prev}`);
}
// Reset start for the next range
if (i < nums.length) {
start = nums[i];
}
}
prev = nums[i - 1];
}
return result.join(',');
}
rust_kv_get — k=v parsing: first '=' splits, last duplicate wins · 4 examplescompile error: implemented for `&str`
= note: required for `&&str` to implement `PartialEq<&str>`
help: consider dereferencing here
|
5 | .find(|(k, _)| *k == key)
| fn kv_get(s: &str, key: &str) -> String {
s.split(';')
.filter_map(|entry| entry.split_once('='))
.find(|(k, _)| k == key)
.map(|(_, v)| v.to_string())
.unwrap_or_default()
}
compile error: implemented for `&str`
= note: required for `&&str` to implement `PartialEq<&str>`
help: consider dereferencing here
|
5 | .find(|(k, _)| *k == key)
| fn kv_get(s: &str, key: &str) -> String {
s.split(';')
.filter_map(|entry| entry.split_once('='))
.find(|(k, _)| k == key)
.map(|(_, v)| v.to_string())
.unwrap_or_default()
}
compile error: implemented for `&str`
= note: required for `&&str` to implement `PartialEq<&str>`
help: consider dereferencing here
|
5 | .find(|(k, _)| *k == key)
| fn kv_get(s: &str, key: &str) -> String {
s.split(';')
.filter_map(|entry| entry.split_once('='))
.find(|(k, _)| k == key)
.map(|(_, v)| v.to_string())
.unwrap_or_default()
}
compile error: implemented for `&str`
= note: required for `&&str` to implement `PartialEq<&str>`
help: consider dereferencing here
|
5 | .find(|(k, _)| *k == key)
| fn kv_get(s: &str, key: &str) -> String {
s.split(';')
.filter_map(|entry| entry.split_once('='))
.find(|(k, _)| k == key)
.map(|(_, v)| v.to_string())
.unwrap_or_default()
}
bash_sort_versions — Numeric per-component version sort · 4 examplesstdin "1.10.0⏎1.9.0⏎1.2.10⏎" → "1.10.0 1.9.0 1.2.10", expected "1.2.10\n1.9.0\n1.10.0"#!/bin/bash
# Read all input into an array
mapfile -t versions
# Sort using version comparison
# Using a custom comparison function with sort -t. -k
# We'll pad each component to a fixed width (e.g., 10 digits) for proper numeric sorting
sort -t. -k1,1n -k2,2n -k3,3n -k4,4n -k5,5n -k6,6n -k7,7n -k8,8n -k9,9n -k10,10n <<< "${versions[*]}" | \
while IFS= read -r line; do
# Remove leading zeros from each component for display (optional, but cleaner)
# But since input is already in standard format, we just output as-is
echo "$line"
done
# Alternative approach: use sort with version sort if available (GNU sort)
# If gsort is available, we can use: gsort -V
# But the problem asks for a script that works with standard tools, so we use numeric sort per component
# More robust version that handles variable number of components:
# Read lines, pad each component to fixed width, sort, then strip padding
# But the above approach works for typical 3-component versions
# Actually, let's implement a more robust solution using printf padding
# This version handles any number of components up to 10
# Read input into array
mapfile -t versions
# Create a temporary file for sorting
tmpfile=$(mktemp)
trap "rm -f $tmpfile" EXIT
for version in "${versions[@]}"; do
# Split version into components
IFS='.' read -ra parts <<< "$version"
# Pad each part to 10 digits with leading zeros
padded=""
for part in "${parts[@]}"; do
# Handle empty parts and ensure numeric
if [[ "$part" =~ ^[0-9]+$ ]]; then
padded+=$(printf "%010d" "$part").
… (truncated)stdin "1.10.0⏎1.9.0⏎1.2.10⏎" → "1.10.0\n1.2.10\n1.9.0", expected "1.2.10\n1.9.0\n1.10.0"#!/bin/bash
# Read all input into an array
mapfile -t versions
# Sort using a custom comparison: compare each dot-separated part numerically
IFS=$'\n' sorted=($(for v in "${versions[@]}"; do
# Convert version to a sortable string by padding each numeric part
# Use printf to pad each component to a fixed width (e.g., 10 digits)
printf "%010d" $(echo "$v" | tr '.' ' ')
printf "\t%s\n" "$v"
done | sort -n | cut -f2-))
# Output sorted versions
printf "%s\n" "${sorted[@]}"
stdin "1.10.0⏎1.9.0⏎1.2.10⏎" → "", expected "1.2.10\n1.9.0\n1.10.0"#!/bin/bash # Sort version strings numerically, handling dot-separated components. # Input: Version strings, one per line on stdin. # Output: Sorted version strings, one per line on stdout. sort -V <<< "$(<&0)"
stdin "1.10.0⏎1.9.0⏎1.2.10⏎" → "", expected "1.2.10\n1.9.0\n1.10.0"#!/bin/bash # Sort version strings numerically, handling dot-separated components. # Input: Version strings, one per line on stdin. # Output: Sorted version strings, one per line on stdout. sort -V <<< "$(<&0)"
tokenize_expr — Tokenise arithmetic, None on invalid input · 4 examplestokenize_expr("1+2") → null, expected [1, "+", 2]import re
def tokenize_expr(s: str) -> list | None:
# Regex to find all potential tokens: numbers, operators, or parentheses
# We use a pattern that captures sequences of digits/dots OR single characters
pattern = r'(\d+\.?\d*|\d*\.\d+|[+\-*/()])'
# First, check for invalid characters by removing valid ones and whitespace
# Valid chars: 0-9, '.', '+', '-', '*', '/', '(', ')', and whitespace
valid_chars_pattern = r'[0-9.+\-*/()\s]'
if not re.fullmatch(valid_chars_pattern + r'*', s):
return None
tokens = []
# We use finditer to get all matches of our token pattern
# However, we must ensure that the sum of lengths of tokens and whitespace equals len(s)
# To handle the "whitespace separates tokens" requirement strictly:
# We can split by whitespace first, but operators might not be separated by spaces.
# So we iterate through the string manually or use a regex that captures everything.
# Let's use a scanner approach to ensure no invalid characters are hidden in "unmatched" parts
scanner = re.finditer(r'(\d+\.\d+|\d+\.|\.\d+|\d+|[+\-*/()])|(\s+)|(.)', s)
for match in scanner:
num, space, invalid = match.groups()
if invalid:
return None
if space:
continue
if num:
# Check for multiple dots (though regex handles most, we validate logic)
if num.count('.') > 1:
return None
# Convert to float if '.' exists, else int
tokens.append(float(num) if '.' in num else int(num))
if
… (truncated)hit the token cap mid-answer — SyntaxError: expected ':' (<model>, line 54)def tokenize_expr(s: str) -> list | None:
"""
Tokenise an arithmetic expression.
* Numbers are non‑negative integers or decimals (e.g. '3', '2.5').
Integers become ``int`` objects, decimals become ``float``.
* Operators '+', '-', '*', '/' and parentheses '(' , ')' are single‑character
string tokens.
* Whitespace separates tokens but is ignored.
* A '-' is **always** an operator; it never belongs to a number.
* Return ``None`` if any illegal character appears or a numeric literal has
more than one decimal point, or if a decimal lacks digits on either side.
"""
tokens = []
i, n = 0, len(s)
while i < n:
ch = s[i]
# Skip whitespace
if ch.isspace():
i += 1
continue
# Number (int or float)
if ch.isdigit() or ch == '.':
start = i
dot_cnt = 0
while i < n and (s[i].isdigit() or s[i] == '.'):
if s[i] == '.':
dot_cnt += 1
if dot_cnt > 1:
return None # more than one decimal point
i += 1
num_str = s[start:i]
# Reject a lone '.' or missing digits around the decimal point
if num_str == '.' or ('.' in num_str and (num_str.startswith('.') or num_str.endswith('.'))):
return None
# Convert to appropriate numeric type
if '.' in num_str:
tokens.append(float(num_str))
else:
tokens.append(int(num_str))
… (truncated)hit the token cap mid-answer — SyntaxError: expected ':' (<model>, line 54)def tokenize_expr(s: str) -> list | None:
"""
Tokenise an arithmetic expression.
* Numbers are non‑negative integers or decimals (e.g. '3', '2.5').
Integers become ``int`` objects, decimals become ``float``.
* Operators '+', '-', '*', '/' and parentheses '(' , ')' are single‑character
string tokens.
* Whitespace separates tokens but is ignored.
* A '-' is **always** an operator; it never belongs to a number.
* Return ``None`` if any illegal character appears or a numeric literal has
more than one decimal point, or if a decimal lacks digits on either side.
"""
tokens = []
i, n = 0, len(s)
while i < n:
ch = s[i]
# Skip whitespace
if ch.isspace():
i += 1
continue
# Number (int or float)
if ch.isdigit() or ch == '.':
start = i
dot_cnt = 0
while i < n and (s[i].isdigit() or s[i] == '.'):
if s[i] == '.':
dot_cnt += 1
if dot_cnt > 1:
return None # more than one decimal point
i += 1
num_str = s[start:i]
# Reject a lone '.' or missing digits around the decimal point
if num_str == '.' or ('.' in num_str and (num_str.startswith('.') or num_str.endswith('.'))):
return None
# Convert to appropriate numeric type
if '.' in num_str:
tokens.append(float(num_str))
else:
tokens.append(int(num_str))
… (truncated)tokenize_expr("1.2.3") → null, expected nullimport re
def tokenize_expr(s: str) -> list | None:
tokens = []
for match in re.finditer(r'\s*([+-/*()]|\d+(?:\.\d+)?(?:(?=\s*[+-/*()]))?)', s):
token = match.group().strip()
if not token or not re.fullmatch(r'[+-/*()]|(\d+\.?\d*)', token):
return None
elif '.' in token:
tokens.append(float(token))
else:
tokens.append(int(token) if token.isdigit() else token)
# Check for consecutive dots and extra characters
if any(re.search(r'\.\.', t) or (t.replace('.','').replace('-','').isdigit() == False and not re.match('^[+-/*()]$',t)) for t in [str(t) for t in tokens]):
return None
else:
return tokens
pluck — Column from associative rows — null vs missing key · 4 examplespluck([{"a": 1}, {"b": 9}, {"a": 3}], "a") → [1, null, 3], expected [1, 3]function pluck(array $rows, string $key): array {
return array_map(
fn($row) => array_key_exists($key, $row) ? $row[$key] : null,
$rows
);
}
pluck([{"a": 1}, {"b": 9}, {"a": 3}], "a") → [1, null, 3], expected [1, 3]function pluck(array $rows, string $key): array {
return array_map(
fn($row) => array_key_exists($key, $row) ? $row[$key] : null,
$rows
);
}
pluck([{"a": 1}, {"b": 9}, {"a": 3}], "a") → [1, null, 3], expected [1, 3]function pluck(array $rows, string $key): array {
return array_map(
fn($row) => array_key_exists($key, $row) ? $row[$key] : null,
$rows
);
}
pluck([{"a": 1}, {"b": 9}, {"a": 3}], "a") → [1, null, 3], expected [1, 3]function pluck(array $rows, string $key): array {
return array_map(
fn($row) => array_key_exists($key, $row) ? $row[$key] : null,
$rows
);
}
topo_lex — Smallest topological order, None on cycle · 4 examplestopo_lex(3, [[1, 0]]) → [1, 2, 0], expected [1, 0, 2]def topo_lex(n: int, edges: list) -> list | None:
from collections import deque
# Build adjacency list and in-degree count
adj = [[] for _ in range(n)]
in_degree = [0] * n
for u, v in edges:
if u != v: # Ignore self-loops (they create cycles)
adj[u].append(v)
in_degree[v] += 1
# Initialize queue with nodes having zero in-degree
q = deque()
for i in range(n):
if in_degree[i] == 0:
q.append(i)
topo_order = []
while q:
# To get lexicographically smallest, process nodes in order
u = q.popleft()
topo_order.append(u)
for v in adj[u]:
in_degree[v] -= 1
if in_degree[v] == 0:
q.append(v)
# Check if topological sort is possible (no cycle)
if len(topo_order) != n:
return None
return topo_order
topo_lex(3, [[1, 0]]) → [1, 2, 0], expected [1, 0, 2]def topo_lex(n: int, edges: list) -> list | None:
from collections import deque
# Build adjacency list and in-degree count
adj = [[] for _ in range(n)]
in_degree = [0] * n
for u, v in edges:
if u != v: # Ignore self-loops (they create cycles)
adj[u].append(v)
in_degree[v] += 1
# Initialize queue with nodes having zero in-degree
q = deque()
for i in range(n):
if in_degree[i] == 0:
q.append(i)
topo_order = []
while q:
# To get lexicographically smallest, process nodes in order
u = q.popleft()
topo_order.append(u)
for v in adj[u]:
in_degree[v] -= 1
if in_degree[v] == 0:
q.append(v)
# Check if topological sort is possible (no cycle)
if len(topo_order) != n:
return None
return topo_order
topo_lex(3, [[1, 0]]) → [1, 2, 0], expected [1, 0, 2]def topo_lex(n: int, edges: list) -> list | None:
from collections import defaultdict, deque
# Build adjacency list and in-degree count
adj = defaultdict(list)
in_degree = [0] * n
for u, v in edges:
if u == v or u < 0 or u >= n or v < 0 or v >= n:
continue # Skip invalid edges (though problem says nodes are 0..n-1)
adj[u].append(v)
in_degree[v] += 1
# Initialize queue with nodes having zero in-degree, sorted for lex order
q = deque()
for i in range(n):
if in_degree[i] == 0:
q.append(i)
topo_order = []
while q:
# Process nodes in lexicographical order
u = q.popleft()
topo_order.append(u)
for v in sorted(adj[u]):
in_degree[v] -= 1
if in_degree[v] == 0:
q.append(v)
if len(topo_order) != n:
return None # Cycle detected
return topo_order
topo_lex(3, [[1, 0]]) → [1, 2, 0], expected [1, 0, 2]def topo_lex(n: int, edges: list) -> list | None:
from collections import deque
# Build adjacency list and in-degree count
adj = [[] for _ in range(n)]
in_degree = [0] * n
for u, v in edges:
if u != v: # Ignore self-loops as they don't affect topological order
adj[u].append(v)
in_degree[v] += 1
# Initialize queue with nodes having zero in-degree, sorted lexicographically
q = deque()
for i in range(n):
if in_degree[i] == 0:
q.append(i)
result = []
while q:
# Process nodes in lexicographical order
u = q.popleft()
result.append(u)
for v in sorted(adj[u]):
in_degree[v] -= 1
if in_degree[v] == 0:
q.append(v)
return result if len(result) == n else None
clamp_mul — Saturating int multiplication · 4 examplescompile error: the current context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(9,19): error CS0103: The name 'Math' does not exipublic static class Sol
{
public static int ClampMul(int a, int b)
{
if (a == 0 || b == 0) return 0;
bool negative = (a < 0) ^ (b < 0);
long la = Math.Abs((long)a);
long lb = Math.Abs((long)b);
long product = la * lb;
if (negative)
{
if (product > (long)(uint.MaxValue >> 1) + 1) // overflow below int.MinValue
return int.MinValue;
return -(int)product;
}
else
{
if (product > int.MaxValue)
return int.MaxValue;
return (int)product;
}
}
}
compile error: the current context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(9,19): error CS0103: The name 'Math' does not exipublic static class Sol
{
public static int ClampMul(int a, int b)
{
if (a == 0 || b == 0) return 0;
bool negative = (a < 0) ^ (b < 0);
long la = Math.Abs((long)a);
long lb = Math.Abs((long)b);
long product = la * lb;
if (negative)
{
if (product > (long)(uint.MaxValue >> 1) + 1) // overflow below int.MinValue
return int.MinValue;
return -(int)product;
}
else
{
if (product > int.MaxValue)
return int.MaxValue;
return (int)product;
}
}
}
compile error: the current context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(9,19): error CS0103: The name 'Math' does not exipublic static class Sol
{
public static int ClampMul(int a, int b)
{
if (a == 0 || b == 0) return 0;
bool negative = (a < 0) ^ (b < 0);
long la = Math.Abs((long)a);
long lb = Math.Abs((long)b);
long product = la * lb;
if (negative)
{
if (product > (long)(uint.MaxValue >> 1) + 1) // overflow below int.MinValue
return int.MinValue;
return -(int)product;
}
else
{
if (product > int.MaxValue)
return int.MaxValue;
return (int)product;
}
}
}
compile error: the current context [/home/crimson/.cache/ai_proxy_cs/csrunner.csproj]
/home/crimson/.cache/ai_proxy_cs/Program.cs(9,19): error CS0103: The name 'Math' does not exipublic static class Sol
{
public static int ClampMul(int a, int b)
{
if (a == 0 || b == 0) return 0;
bool negative = (a < 0) ^ (b < 0);
long la = Math.Abs((long)a);
long lb = Math.Abs((long)b);
long product = la * lb;
if (negative)
{
if (product > (long)(uint.MaxValue >> 1) + 1) // overflow below int.MinValue
return int.MinValue;
return -(int)product;
}
else
{
if (product > int.MaxValue)
return int.MaxValue;
return (int)product;
}
}
}
lru_ops — LRU cache with eviction order · 4 exampleslru_ops(2, [["put", 1, 1], ["put", 2, 2], ["put", 1, 10], ["put", 3, 3], ["get", 2], ["get", 1]]) → null, expected [-1, 10]def lru_ops(capacity: int, ops: list) -> list:
"""Simulates an LRU cache and returns results of 'get' operations.
Args:
capacity: The maximum capacity of the cache.
ops: A list of operations to perform on the cache. Each operation is a list
of the form ['put', key, value] or ['get', key].
Returns:
A list of results of the 'get' operations only, using -1 for a miss.
"""
cache = {}
usage = [] # List to track usage order (most recent last)
results = []
for op in ops:
if op[0] == 'put':
key = op[1]
value = op[2]
if key in cache:
cache[key] = value
usage.remove(key) # Move to end (most recently used)
else:
if len(cache) >= capacity:
lru_key = usage.pop(0) # Evict least recently used
del cache[lru_key]
cache[key] = value
usage.append(key) # Add to end (most recently used)
elif op[0] == 'get':
key = op[1]
if key in cache:
usage.remove(key) # Move to end (most recently used)
usage.append(key)
results.append(cache[key])
else:
results.append(-1)
return results
lru_ops(2, [["put", 1, 1], ["put", 2, 2], ["put", 1, 10], ["put", 3, 3], ["get", 2], ["get", 1]]) → null, expected [-1, 10]def lru_ops(capacity: int, ops: list) -> list:
"""Simulates an LRU cache and returns results of 'get' operations.
Args:
capacity: The maximum capacity of the cache.
ops: A list of operations to perform on the cache. Each operation is a list
of the form ['put', key, value] or ['get', key].
Returns:
A list of results of the 'get' operations only, using -1 for a miss.
"""
cache = {}
usage = [] # List to track usage order (most recent last)
results = []
for op in ops:
if op[0] == 'put':
key = op[1]
value = op[2]
if key in cache:
cache[key] = value
usage.remove(key) # Move to end (most recently used)
else:
if len(cache) >= capacity:
lru_key = usage.pop(0) # Evict least recently used
del cache[lru_key]
cache[key] = value
usage.append(key) # Add to end (most recently used)
elif op[0] == 'get':
key = op[1]
if key in cache:
usage.remove(key) # Move to end (most recently used)
results.append(cache[key])
else:
results.append(-1)
return results
SyntaxError: invalid syntax (<model>, line 1)```python code
def lru_ops(capacity: int, ops: list) -> list:
cache = {}
usage_order = []
def update_usage(key):
if key in usage_order:
usage_order.remove(key)
usage_order.append(key)
results = []
for op in ops:
if op[0] == 'put':
key, value = op[1], op[2]
if key in cache:
cache[key] = value
update_usage(key)
else:
if len(cache) >= capacity:
lru_key = usage_order.pop(0)
del cache[lru_key]
cache[key] = value
update_usage(key)
elif op[0] == 'get':
key = op[1]
if key in cache:
update_usage(key)
results.append(cache[key])
else:
results.append(-1)
return results
```SyntaxError: invalid syntax (<model>, line 1)```python code
def lru_ops(capacity: int, ops: list) -> list:
cache = {}
usage_order = []
def update_usage(key):
if key in usage_order:
usage_order.remove(key)
usage_order.append(key)
results = []
for op in ops:
if op[0] == 'put':
key, value = op[1], op[2]
if key in cache:
cache[key] = value
update_usage(key)
else:
if len(cache) >= capacity:
lru_key = usage_order.pop(0)
del cache[lru_key]
cache[key] = value
update_usage(key)
elif op[0] == 'get':
key = op[1]
if key in cache:
update_usage(key)
results.append(cache[key])
else:
results.append(-1)
return results
```expand_ranges — Expand "1-3,7" specs with validation · 4 exampleshit the token cap mid-answer — expand_ranges("1-3,7") → null, expected [1, 2, 3, 7]def expand_ranges(s: str) -> list | None:
"""
Expand a comma‑separated range specification into a sorted list of unique ints.
Examples
--------
>>> expand_ranges("1-3, 7 ,10 -12")
[1, 2, 3, 7, 10, 11, 12]
Returns None for any malformed input:
* empty string or only whitespace
* empty part (e.g. "1,,3" or trailing comma)
* non‑numeric tokens
* reversed range like "5-3"
* more than one hyphen in a part
"""
if not isinstance(s, str):
return None
s = s.strip()
if not s:
return None
parts = s.split(',')
intervals: list[tuple[int, int]] = []
for raw_part in parts:
part = raw_part.strip()
if not part: # empty part
return None
if '-' in part:
subparts = part.split('-')
hit the token cap mid-answer — SyntaxError: unterminated string literal (detected at line 86) (<model>, line 86)import re
def expand_ranges(s: str) -> list | None:
"""
Expands a spec like "1-3,7,10-12" into a sorted list of unique ints.
Whitespace around numbers, hyphens and commas is allowed. Overlapping ranges merge.
Invalid input — empty string, empty part, non-numeric, or a reversed range like "5-3" — returns None.
"""
if not s:
return None
# Split by comma, allowing for surrounding whitespace
parts = [part.strip() for part in s.split(',')]
if any(not part for part in parts):
return None
ranges = []
for part in parts:
part = part.strip()
if not part:
# This case should be caught by the initial check, but good for robustness
return None
if '-' in part:
try:
start_str, end_str = map(str.strip, part.split('-'))
start = int(start_str)
end = int(end_str)
if start > end:
return None # Reversed range
ranges.append((start, end))
except ValueError:
return None # Non-numeric part in range
else:
try:
num = int(part)
ranges.append((num, num))
except ValueError:
return None # Non-numeric standalone part
if not ranges:
# Should be caught by the initial check, but handles case where s was just commas/spaces resulting in empty parts list if logic above failed slightly
return None
# 1. Sort ranges ba
… (truncated)hit the token cap mid-answer — SyntaxError: '(' was never closed (<model>, line 87)import re
def expand_ranges(s: str) -> list | None:
"""
Expands a spec like "1-3,7,10-12" into a sorted list of unique ints.
Whitespace around numbers, hyphens and commas is allowed. Overlapping ranges merge.
Invalid input — empty string, empty part, non-numeric, or a reversed range like "5-3" — returns None.
"""
if not s:
return None
# Normalize whitespace and split by comma
parts = [part.strip() for part in s.split(',')]
if any(not part for part in parts):
return None
ranges = []
for part in parts:
part = part.strip()
if not part:
return None
if '-' in part:
try:
start_str, end_str = map(str.strip, part.split('-'))
start = int(start_str)
end = int(end_str)
if start > end:
return None # Reversed range
ranges.append((start, end))
except ValueError:
return None # Non-numeric part in range
else:
try:
num = int(part)
ranges.append((num, num))
except ValueError:
return None # Non-numeric standalone part
if not ranges:
# Should be caught by the initial empty string check, but for safety
return None
# Sort ranges based on start point
ranges.sort()
# Merge overlapping ranges and expand to list of unique integers
merged_ranges = []
if ranges:
current_start, current_end = ran
… (truncated)expand_ranges("1-3,7") → null, expected [1, 2, 3, 7]def expand_ranges(s: str) -> list | None:
if not s.strip():
return None
ranges = [part.split('-') for part in s.replace(',', '-').split()]
result = set()
for r in ranges:
if len(r) != 2 or not all(x.strip().isdigit() for x in r):
return None
start, end = int(r[0]), int(r[1])
if start > end:
return None
result.update(range(start, end + 1))
return sorted(list(result))
path_norm — Normalise a POSIX path with . and .. · 4 examplespath_norm("/../") → "/..", expected "/"def path_norm(path: str) -> str:
if not path:
return '/'
# Split path into components
components = []
current = ''
for char in path:
if char == '/':
if current:
components.append(current)
current = ''
else:
current += char
# Add the last component if it exists
if current:
components.append(current)
# Process components
stack = []
for component in components:
if component == '.' or component == '':
# Skip current directory references and empty components
continue
elif component == '..':
# Go up one level if possible
if stack and stack[-1] != '..':
stack.pop()
else:
# Stay at root if already at root
stack.append(component)
else:
stack.append(component)
# Handle the case where we're at the root
# If we have a .. at the beginning, we should keep it
result = '/' + '/'.join(stack)
# If result is just '/', return it
if result == '/':
return '/'
# Remove trailing slash
if result.endswith('/'):
result = result[:-1]
return result
path_norm("/../") → "/..", expected "/"def path_norm(path: str) -> str:
if not path:
return '/'
# Split path into components
components = []
current = ''
for char in path:
if char == '/':
if current:
components.append(current)
current = ''
else:
current += char
# Add the last component if it exists
if current:
components.append(current)
# Process components
stack = []
for component in components:
if component == '.' or component == '':
# Skip current directory references and empty components
continue
elif component == '..':
# Go up one level if possible
if stack and stack[-1] != '..':
stack.pop()
elif not stack or stack[-1] == '..':
# Stay at root or add .. if already at root
stack.append('..')
else:
stack.append(component)
# Handle special case where we're at the root but have .. components
# This is handled by the logic above, but let's make sure
result = '/' + '/'.join(stack) if stack else '/'
return result
path_norm("/../") → "/..", expected "/"def path_norm(path: str) -> str:
if not path:
return '/'
# Split path by '/' and filter out empty strings
parts = [part for part in path.split('/') if part]
# Process parts to handle '.' and '..'
stack = []
for part in parts:
if part == '.':
continue
elif part == '..':
if stack and stack[-1] != '..':
stack.pop()
elif not stack or stack[-1] == '..':
stack.append('..')
else:
stack.append(part)
# Handle case where we're at root and have '..'
# This ensures that '..' at the root stays at the root
if not stack and path.startswith('/'):
return '/'
# Reconstruct path
result = '/' + '/'.join(stack)
# Remove trailing slash if it exists (but keep root '/')
if result != '/' and result.endswith('/'):
result = result[:-1]
return result
path_norm("/../") → "/..", expected "/"def path_norm(path: str) -> str:
if not path:
return '/'
# Split path into components
components = []
current = ''
for char in path:
if char == '/':
if current:
components.append(current)
current = ''
else:
# Handle multiple consecutive slashes
pass
else:
current += char
# Add the last component if it exists
if current:
components.append(current)
# Process components
stack = []
for component in components:
if component == '.' or component == '':
# Skip current directory references and empty components
continue
elif component == '..':
# Go up one level if possible
if stack and stack[-1] != '..':
stack.pop()
else:
# Stay at root if already at root
stack.append(component)
else:
stack.append(component)
# Special case: if stack is empty but we started with a root, return root
if not stack and path.startswith('/'):
return '/'
# Reconstruct the path
result = '/' + '/'.join(stack)
# Remove trailing slash (but keep root)
if len(result) > 1 and result.endswith('/'):
result = result[:-1]
return result
json_pointer — Resolve an RFC 6901 JSON Pointer · 4 examplesjson_pointer({"a/b": 7}, "/a~1b") → null, expected 7import json
def json_pointer(doc, pointer: str):
"""
Resolves an RFC 6901 JSON Pointer against doc, returning the referenced value or None if it does not resolve.
Args:
doc: The document (Python object) to query.
pointer: The JSON Pointer string.
Returns:
The referenced value, or None if the pointer is invalid or out of bounds.
"""
if not pointer:
return doc
# 1. Decode the pointer components
def decode_token(token):
# Handle escaped characters first (RFC 6901 specifies '~1' for '/' and '~0' for '~')
if token == "~1":
return "/"
elif token == "~0":
return "~"
else:
return token
# The pointer is split by '/', but we must handle the escaped slashes first.
# A simple split('/') will break on encoded '/' ('~1').
# We need a custom splitting mechanism or careful replacement/splitting.
# Since '~1' decodes to '/', and we are splitting by '/', we can temporarily
# replace all literal '/' with a unique placeholder, then decode the escaped ones,
# and finally split by the original separator logic (which is now safe).
# A simpler approach based on standard library behavior for JSON Pointer decoding:
# The pointer string needs to be decoded according to RFC 6901 rules.
# We can use a regex or iterative replacement, but since we are implementing it manually:
tokens = []
current_token = ""
i = 0
while i < len(pointer):
char = pointer[i]
if char == '/':
# If the current t
… (truncated)hit the token cap mid-answer — SyntaxError: invalid syntax (<model>, line 98)import re
def json_pointer(doc, pointer: str):
"""
Resolves an RFC 6901 JSON Pointer against doc, returning the referenced value or None if it does not resolve.
Args:
doc: The document (Python object) to query.
pointer: The JSON Pointer string.
Returns:
The referenced value, or None if the pointer is invalid or out of bounds.
"""
if not pointer:
return doc
# 1. Decode the pointer according to RFC 6901 rules for escaped characters
def decode_token(token):
# '~1' decodes to '/' and '~0' decodes to '~'
# The rule states '~1' must be decoded BEFORE '~0'.
decoded = token.replace('~1', '/').replace('~0', '~')
return decoded
# Split the pointer by '/', but we need a custom split/decode process because
# the components themselves might contain escaped slashes or tildes.
# A simpler approach is to first decode the whole string, then split by '/'
# However, standard JSON Pointer decoding requires careful handling of escapes *within* tokens.
# Let's use a regex-based iterative decoding/splitting process that respects escaping:
tokens = []
current_token = ""
i = 0
while i < len(pointer):
char = pointer[i]
if char == '/':
if current_token:
# Decode the accumulated token before appending it
decoded_token = decode_token(current_token)
tokens.append(decoded_token)
current_token = ""
else:
# Handle leading/consecutive slashes if th
… (truncated)hit the token cap mid-answer — SyntaxError: invalid syntax (<model>, line 98)import re
def json_pointer(doc, pointer: str):
"""
Resolves an RFC 6901 JSON Pointer against doc, returning the referenced value or None if it does not resolve.
Args:
doc: The document (Python object) to query.
pointer: The JSON Pointer string.
Returns:
The referenced value, or None if the pointer is invalid or out of bounds.
"""
if not pointer:
return doc
# 1. Decode the pointer according to RFC 6901 rules for escaped characters
def decode_token(token):
# '~1' decodes to '/' and '~0' decodes to '~'
# The rule states '~1' must be decoded BEFORE '~0'.
decoded = token.replace('~1', '/').replace('~0', '~')
return decoded
# Split the pointer by '/', but we need a custom split/decode process because
# the components themselves might contain escaped slashes or tildes.
# A simpler approach is to first decode the whole string, then split by '/'
# However, standard JSON Pointer decoding requires careful handling of escapes *within* tokens.
# Let's use a regex-based iterative decoding/splitting process that respects escaping:
tokens = []
current_token = ""
i = 0
while i < len(pointer):
char = pointer[i]
if char == '/':
if current_token:
# Decode the accumulated token before appending it
decoded_token = decode_token(current_token)
tokens.append(decoded_token)
current_token = ""
else:
# Handle leading/consecutive slashes if th
… (truncated)hit the token cap mid-answer — SyntaxError: invalid syntax (<model>, line 97)import re
def json_pointer(doc, pointer: str):
"""
Resolves an RFC 6901 JSON Pointer against doc, returning the referenced value or None if it does not resolve.
Args:
doc: The document (Python object) to query.
pointer: The JSON Pointer string.
Returns:
The referenced value, or None if the pointer is invalid or out of bounds.
"""
if not pointer:
return doc
# 1. Decode the pointer according to RFC 6901 rules for escaped characters
def decode_token(token):
# '~1' decodes to '/' and '~0' decodes to '~'
# The rule states '~1' must be decoded BEFORE '~0'.
decoded = token.replace('~1', '/').replace('~0', '~')
return decoded
# Split the pointer by '/', but we need a custom split/decode process because
# the components themselves might contain escaped slashes or tildes.
# A simpler approach is to first decode the whole string, then split by '/'
# However, standard JSON Pointer decoding requires careful handling of escapes *within* tokens.
# Let's use a regex-based iterative decoding/splitting process that respects escaping:
tokens = []
current_token = ""
i = 0
while i < len(pointer):
char = pointer[i]
if char == '/':
if current_token:
# Decode the accumulated token before appending it
decoded_token = decode_token(current_token)
tokens.append(decoded_token)
current_token = ""
else:
# Handle leading/consecutive slashes if th
… (truncated)clamp_add — Saturating int addition — overflow trap · 4 examplesclamp_add(100, 200) → "-2147483648", expected 300int clamp_add(int a, int b) {
if (a > INT_MAX - b) return INT_MAX;
if (a < INT_MIN - b) return INT_MIN;
return a + b;
}
clamp_add(100, 200) → "-2147483648", expected 300int clamp_add(int a, int b) {
if (a > INT_MAX - b) return INT_MAX;
if (a < INT_MIN - b) return INT_MIN;
return a + b;
}
clamp_add(100, 200) → "-2147483648", expected 300int clamp_add(int a, int b) {
if (a > INT_MAX - b) return INT_MAX;
if (a < INT_MIN - b) return INT_MIN;
return a + b;
}
clamp_add(100, 200) → "-2147483648", expected 300int clamp_add(int a, int b) {
if (a > INT_MAX - b) return INT_MAX;
if (a < INT_MIN - b) return INT_MIN;
return a + b;
}
ring_buffer — Ring buffer that overwrites the oldest on overflow · 4 examplesring(2, [["push", 1], ["push", 2], ["push", 3], ["pop"], ["pop"], ["pop"]]) → [3, 2, null], expected [2, 3, null]def ring(capacity: int, ops: list) -> list:
buffer = [None] * capacity
head = 0 # index of the oldest element
count = 0 # number of elements currently in the buffer
results = []
for op in ops:
if op[0] == 'push':
value = op[1]
if count == capacity:
# Buffer is full, overwrite the oldest element
head = (head + 1) % capacity
else:
count += 1
buffer[head] = value
elif op[0] == 'pop':
if count == 0:
results.append(None)
else:
result = buffer[head]
head = (head + 1) % capacity
count -= 1
results.append(result)
return results
IndentationError: unexpected indent (<model>, line 1) def ring(capacity, ops):
buffer = []
results = []
for op in ops:
if op[0] == 'push':
if len(buffer) < capacity:
buffer.append(op[1])
else:
buffer.pop(0) # Remove the oldest element to make space
buffer.append(op[1])
elif op[0] == 'pop':
results.append(buffer.pop(0) if len(buffer) > 0 else None)
return results IndentationError: unexpected indent (<model>, line 1) def ring(capacity, ops):
buffer = []
results = []
for op in ops:
if op[0] == 'push':
if len(buffer) < capacity:
buffer.append(op[1])
else:
buffer.pop(0) # Remove the oldest element to make space
buffer.append(op[1])
elif op[0] == 'pop':
results.append(buffer.pop(0) if len(buffer) > 0 else None)
return results IndentationError: unexpected indent (<model>, line 1) def ring(capacity, ops):
buffer = []
results = []
for op in ops:
if op[0] == 'push':
if len(buffer) < capacity:
buffer.append(op[1])
else:
buffer.pop(0) # Remove the oldest element from the front of the list
buffer.append(op[1])
elif op[0] == 'pop':
if len(buffer) > 0:
results.append(buffer.pop(0)) # Remove and return the oldest element from the front of the list
else:
results.append(None)
return results csv_escape — RFC 4180 CSV field quoting · 4 examplescsv_escape("say \"hi\"") → "\"say \"hi\"\"", expected "\"say \"\"hi\"\"\""#include <string>
std::string csv_escape(std::string field) {
bool needs_quoting = false;
for (char c : field) {
if (c == ',' || c == '"') {
needs_quoting = true;
break;
}
}
if (!needs_quoting) {
return field;
}
std::string escaped;
escaped.reserve(field.size() + 2);
escaped += '"';
for (char c : field) {
if (c == '"') {
escaped += "\"";
} else {
escaped += c;
}
}
escaped += '"';
return escaped;
}
compile error: = char; _Traits = std::char_traits<char>; _Alloc = std::allocator<char>; size_type = long unsigned int]'
2713 | find(_CharT __c, size_type __pos = 0) const _#include <string>
std::string csv_escape(const std::string& field) {
if (field.empty()) return "";
size_t comma_pos = field.find(',');
size_t quote_pos = field.find('"');
bool has_comma = (comma_pos != std::string::npos && comma_pos <= field.size() - 1);
bool has_quote = (quote_pos != std::string::npos && quote_pos <= field.size() - 2);
if (has_comma || has_quote) {
size_t start = 0;
std::string escaped;
while (start < field.size()) {
size_t end = field.find(',', start, 1);
if (end == std::string::npos) end = field.size();
size_t quote_end = field.find('"', start + 1);
if (quote_end != std::string::npos && quote_end <= end - 2) {
escaped += '"';
for (size_t i = start + 1; i < quote_end; ++i) {
char c = field[i];
if (c == '"') escaped += '"';
else escaped += c;
}
escaped += '"';
start = quote_end + 1;
} else {
size_t sub_start = end > start ? end : std::string::npos;
for (size_t i = start; i < field.size() && i <= sub_start; ++i) {
char c = field[i];
if (c == '"') escaped += '"';
else escaped += c;
}
start = sub_start + 1;
}
}
return '"' + escaped + '"';
} else {
return field;
}
}
compile error: _cxx11::basic_string<char>::const_iterator; const_iterator = std::__cxx11::basic_string<char>::const_iterator]'
2435 | replace(__const_iterator __i1, __const_#include <string>
std::string csv_escape(const std::string& field) {
if (field.empty()) return "";
size_t comma_pos = field.find(',');
size_t quote_pos = field.find('"');
bool has_comma = (comma_pos != std::string::npos && comma_pos <= field.size() - 1);
bool has_quote = (quote_pos != std::string::npos && quote_pos <= field.size() - 2);
if (has_comma || has_quote) {
size_t start = 0;
while (true) {
size_t end = field.find('=', start + 1, 0);
if (end == std::string::npos) break;
size_t quote_end = field.find('"', end + 2, 0);
if (quote_end != std::string::npos && quote_end < field.size() - 1) {
size_t quote_len = quote_end - end - 1;
for (size_t i = 0; i <= quote_len; ++i) {
char c = field[end + 2 + i];
if (c == '"') {
c = '"';
}
std::string part = field.substr(end, quote_end - end + 1);
size_t new_start = start;
size_t new_end = end + quote_len + 1;
std::string escaped_part = part;
for (size_t i = 0; i < escaped_part.size(); ++i) {
if (escaped_part[i] == '"') {
escaped_part.insert(i, 1, '"');
++i;
}
}
field.replace(new_start, new_end - new_start, escaped_part);
}
} else {
break;
… (truncated)compile error: = char; _Traits = std::char_traits<char>; _Alloc = std::allocator<char>; size_type = long unsigned int]'
2713 | find(_CharT __c, size_type __pos = 0) const _#include <string>
std::string csv_escape(const std::string& field) {
if (field.empty()) return "";
size_t comma_pos = field.find(',');
size_t quote_pos = field.find('"');
bool has_comma = (comma_pos != std::string::npos && comma_pos <= field.size() - 1);
bool has_quote = (quote_pos != std::string::npos && quote_pos <= field.size() - 2);
if (has_comma || has_quote) {
size_t start = 0;
std::string escaped;
while (start < field.size()) {
size_t end = field.find(',', start, 1);
if (end == std::string::npos) end = field.size();
size_t quote_end = field.find('"', start + 1);
if (quote_end != std::string::npos && quote_end <= end - 2) {
escaped += '"';
for (size_t i = start + 1; i < quote_end; ++i) {
char c = field[i];
if (c == '"') escaped += '"';
else escaped += c;
}
escaped += '"';
start = quote_end + 1;
} else {
size_t sub_start = end > start ? end : std::string::npos;
for (size_t i = start; i < field.size() && i <= sub_start; ++i) {
char c = field[i];
if (c == '"') escaped += '"';
else escaped += c;
}
start = sub_start + 1;
}
}
return '"' + escaped + '"';
} else {
return field;
}
}
bash_dedup_lines — First occurrence wins, order preserved (sort -u reorders) · 4 examplesstdin "b⏎a⏎b⏎c⏎a⏎" → "/tmp/bench_bash_z8wq5w0n/task.sh: line 1: your_script.sh: co\u2026", expected "b\na\nc"cat input_file.txt | your_script.sh
stdin "b⏎a⏎b⏎c⏎a⏎" → "apple\\nbanana\\napple\\norange", expected "b\na\nc"echo "apple\nbanana\napple\norange" | {
declare -A seen;
while IFS= read -r line; do
if [[ -z ${seen[$line]} ]]; then
echo "$line";
seen[$line]=1;
fi
done;
}
stdin "b⏎a⏎b⏎c⏎a⏎" → "apple\\nbanana\\napple\\norange", expected "b\na\nc"echo "apple\nbanana\napple\norange" | {
declare -A seen;
while IFS= read -r line; do
if [[ -z ${seen[$line]} ]]; then
echo "$line";
seen[$line]=1;
fi
done;
}
stdin "b⏎a⏎b⏎c⏎a⏎" → "apple\\nbanana\\napple\\norange", expected "b\na\nc"echo "apple\nbanana\napple\norange" | {
declare -A seen;
while IFS= read -r line; do
if [[ -z ${seen[$line]} ]]; then
echo "$line";
seen[$line]=1;
fi
done;
}
count_words — Count words split on spaces and tabs · 4 examplesno code in response Here is the C function to count words in a given string:
```cpp
#include <stdio.h>
#include <ctype.h>
int count_words(const char *s) {
int wordCount = 0;
while (*s != '\0') {
// Skip leading separators (spaces and tabs)
while (*s == ' ' || *s == '\t') {
s++;
}
if (*s != '\0') {
wordCount++;
// Skip the current word
while (*s != ' ' && *s != '\t' && *s != '\0') {
s++;
}
}
}
return wordCount;
}
```
This function iterates through theno code in response Here is the C function to count words in a given string:
```cpp
#include <stdio.h>
#include <ctype.h>
int count_words(const char *s) {
int wordCount = 0;
while (*s != '\0') {
// Skip leading separators (spaces and tabs)
while (*s == ' ' || *s == '\t') {
s++;
}
if (*s != '\0') {
wordCount++;
// Skip the current word
while (*s != ' ' && *s != '\t' && *s != '\0') {
s++;
}
}
}
return wordCount;
}
```
This function iterates through theno code in response Here is the C function to count words in a given string:
```cpp
#include <stdio.h>
#include <ctype.h>
int count_words(const char *s) {
int wordCount = 0;
while (*s != '\0') {
// Skip leading separators (spaces and tabs)
while (*s == ' ' || *s == '\t') {
s++;
}
if (*s != '\0') {
wordCount++;
// Skip the current word
while (*s != ' ' && *s != '\t' && *s != '\0') {
s++;
}
}
}
return wordCount;
}
```
This function iterates through theno code in response Here is the C function to count words in a given string:
```cpp
#include <stdio.h>
#include <ctype.h>
int count_words(const char *s) {
int wordCount = 0;
while (*s != '\0') {
// Skip leading separators (spaces and tabs)
while (*s == ' ' || *s == '\t') {
s++;
}
if (*s != '\0') {
wordCount++;
// Skip the current word
while (*s != ' ' && *s != '\t' && *s != '\0') {
s++;
}
}
}
return wordCount;
}
```
This function iterates through thesql_missing_users — Anti-join where NOT IN meets a NULL · 4 examplesquery returned [], expected [["bob"], ["cy"]]SELECT users.name FROM users WHERE users.id NOT IN (SELECT DISTINCT orders.user_id FROM orders) ORDER BY users.name ASC;
query returned [], expected [["bob"], ["cy"]]SELECT users.name FROM users WHERE users.id NOT IN (SELECT DISTINCT orders.user_id FROM orders) ORDER BY users.name ASC;
query returned [], expected [["bob"], ["cy"]]SELECT users.name FROM users WHERE users.id NOT IN (SELECT DISTINCT orders.user_id FROM orders) ORDER BY users.name ASC;
query returned [], expected [["bob"], ["cy"]]SELECT users.name FROM users WHERE users.id NOT IN (SELECT DISTINCT orders.user_id FROM orders) ORDER BY users.name ASC;
interval_intersect — Intersect two interval lists · 4 examplesinterval_intersect([[0, 2], [5, 10]], [[1, 5], [8, 12]]) → [[1, 2]], expected [[1, 2], [5, 5], [8, 10]]def interval_intersect(a: list, b: list) -> list:
i = j = 0
res = []
while i < len(a) and j < len(b):
start = max(a[i][0], b[j][0])
end = min(a[i][1], b[j][1])
if start <= end:
res.append([start, end])
if a[i][1] > b[j][1]:
i += 1
else:
j += 1
return res
interval_intersect([[0, 2], [5, 10]], [[1, 5], [8, 12]]) → [[1, 2]], expected [[1, 2], [5, 5], [8, 10]]def interval_intersect(a: list, b: list) -> list:
i = j = 0
res = []
while i < len(a) and j < len(b):
start = max(a[i][0], b[j][0])
end = min(a[i][1], b[j][1])
if start <= end:
res.append([start, end])
if a[i][1] > b[j][1]:
i += 1
else:
j += 1
return res
interval_intersect([[0, 2], [5, 10]], [[1, 5], [8, 12]]) → [[1, 2]], expected [[1, 2], [5, 5], [8, 10]]def interval_intersect(a: list, b: list) -> list:
i = j = 0
res = []
while i < len(a) and j < len(b):
start = max(a[i][0], b[j][0])
end = min(a[i][1], b[j][1])
if start <= end:
res.append([start, end])
if a[i][1] > b[j][1]:
i += 1
else:
j += 1
return res
interval_intersect([[0, 2], [5, 10]], [[1, 5], [8, 12]]) → [[1, 2]], expected [[1, 2], [5, 5], [8, 10]]def interval_intersect(a: list, b: list) -> list:
i = j = 0
res = []
while i < len(a) and j < len(b):
start = max(a[i][0], b[j][0])
end = min(a[i][1], b[j][1])
if start <= end:
res.append([start, end])
if a[i][1] > b[j][1]:
i += 1
else:
j += 1
return res
glob_match — Glob matching with ? and * · 4 examplesglob_match("a?c", "ac") → null, expected falsedef glob_match(pattern: str, text: str) -> bool:
def match(text_index, pattern_index):
if text_index == len(text) and pattern_index == len(pattern):
return True # Both strings are fully matched.
elif text_index != len(text) and pattern_index == len(pattern):
return False # Pattern is exhausted but there's still text left.
elif text_index == len(text) and pattern[pattern_index] == '*':
return match(text_index, pattern_index + 1) # Text is exhausted; skip asterisks.
else:
if (pattern[pattern_index] == '?' and text_index < len(text)) or (pattern[pattern_index] == text[text_index]):
return match(text_index + 1, pattern_index + 1) # Matched a character or question mark.
elif pattern[pattern_index] == '*':
return match(text_index, pattern_index + 1) or match(text_index + 1, pattern_index) # Skip the asterisk or skip both.
else:
return False # Mismatch without wildcards.
return match(0, 0)
glob_match("a?c", "ac") → null, expected falsedef glob_match(pattern: str, text: str) -> bool:
def match(text_index, pattern_index):
if text_index == len(text) and pattern_index == len(pattern):
return True # Both strings are fully matched.
elif text_index != len(text) and pattern_index == len(pattern):
return False # Pattern is exhausted but there's still text left.
elif text_index == len(text) and pattern[pattern_index] == '*':
return match(text_index, pattern_index + 1) # Text is exhausted; skip asterisks.
else:
if (pattern[pattern_index] == '?' and text_index < len(text)) or (pattern[pattern_index] == text[text_index]):
return match(text_index + 1, pattern_index + 1) # Matched a character or question mark.
elif pattern[pattern_index] == '*':
return match(text_index, pattern_index + 1) or match(text_index + 1, pattern_index) # Skip the asterisk or skip both.
else:
return False # Mismatch without wildcards.
return match(0, 0)
hit the token cap mid-answer — SyntaxError: unterminated string literal (detected at line 11) (<model>, line 11)We are to write a function that matches a pattern against a text without using re, fnmatch, or glob.
The pattern uses:
'?' matches exactly one character.
'*' matches any sequence (including empty) of characters.
We cannot use the built-in modules, so we have to implement the matching ourselves.
Approach:
We can use recursion or iterative dynamic programming. However, the problem says to return only the function.
Let's design a recursive function that works from the beginning of the pattern and text.
Steps:
1. If the pattern is empty and text is empty -> True.
2. If the pattern is empty but text is not -> False.
3. If the text is empty but pattern is not -> then we check if pattern is all stars? But note: we can have multiple stars.
However, note: the problem says that '*' matches any sequence including empty.
We can do:
Let's define a helper function that takes (i, j) where i is index in pattern and j is index in text.
But the problem says: return only the function. We can do without helper by using recursion with two indices.
Alternatively, we can do iterative from the start.
However, the problem does not specify performance, but we want to be efficient enough.
Idea:
We traverse the pattern and text.
Steps for iterative:
i = 0 (pattern index)
j = 0 (text index)
while i < len(pattern) and j < len(text):
if pattern[i] == '*':
# We have a star: we can skip the star and try to match from the next pattern character with the current text
# But note: star can match empty, so we might want to skip th
… (truncated)hit the token cap mid-answer — SyntaxError: unterminated string literal (detected at line 11) (<model>, line 11)We are to write a function that matches a pattern against a text without using re, fnmatch, or glob.
The pattern uses:
'?' matches exactly one character.
'*' matches any sequence (including empty) of characters.
We cannot use the built-in modules, so we have to implement the matching.
Approach:
We can use recursion or iterative dynamic programming. However, the problem says to return only the function.
Let's design a recursive function that works from the beginning of the pattern and text.
Steps:
1. If the pattern is empty and text is empty -> True.
2. If pattern is empty but text is not -> False.
3. If text is empty but pattern is not -> then we have to check if pattern is all stars? But note: '*' can match empty, so if pattern is non-empty and text is empty, then we can have pattern starting with '*' and then the rest? Actually, we have to consider the entire pattern.
However, note: the problem says "pattern matches the whole of text".
We can do:
Let's define a helper function that takes (i, j) where i is index in pattern, j is index in text.
But the problem says: return only the function. We can do without helper by recursion? But recursion might be deep.
Alternatively, we can do iterative with two pointers.
However, the problem does not specify performance, but we want to be efficient enough.
Idea for iterative:
We traverse the pattern and text.
Steps for iterative:
i = 0 (pattern index)
j = 0 (text index)
while j < len(text):
if pattern[i] == '*':
# We can skip this star and try to match the rest of the patte
… (truncated)theme_vars — Dark-mode token inside a media query · 4 examplesno code in response1. Write CSS implementing a two-theme token: :root defines --ink as #111111; inside an @media (prefers-color-scheme: dark) block, :root redefines --ink as #eeeeee; and body sets color to var(--ink). The trap is scope — the dark value must live inside the media query, not beside it. Return only the CSS in a single ```css code block.
no code in response1. Write CSS implementing a two-theme token: :root defines --ink as #111111; inside an @media (prefers-color-scheme: dark) block, :root redefines --ink as #eeeeee; and body sets color to var(--ink). The trap is scope — the dark value must live inside the media query, not beside it. Return only the CSS in a single ```css code block.
no code in response1. Write CSS implementing a two-theme token: :root defines --ink as #111111; inside an @media (prefers-color-scheme: dark) block, :root redefines --ink as #eeeeee; and body sets color to var(--ink). The trap is scope — the dark value must live inside the media query, not beside it. Return only the CSS in a single ```css code block.
no code in response1. Write CSS implementing a two-theme token: :root defines --ink as #111111; inside an @media (prefers-color-scheme: dark) block, :root redefines --ink as #eeeeee; and body sets color to var(--ink). The trap is scope — the dark value must live inside the media query, not beside it. Return only the CSS in a single ```css code block.
data_table — Revenue table with caption and scoped headers · 4 examplescount of "thead th" → 2, expected 3<table>
<caption>Quarterly revenue</caption>
<thead>
<tr scope="col">Q1</scope>
<th scope="col">Q2</th>
<th scope="col">Q3</th>
</tr>
</thead>
<tbody>
<tr>
<td>Revenue: $50,000</td>
<td>Revenue: $65,000</td>
<td>Revenue: $72,000</td>
</tr>
<tr>
<td>Growth: 10%</td>
<td>Growth: 8%</td>
<td>Growth: 4%</td>
</tr>
</tbody>
</table>
count of "thead th" → 2, expected 3<table>
<caption>Quarterly revenue</caption>
<thead>
<tr scope="col">Q1</scope>
<th scope="col">Q2</th>
<th scope="col">Q3</th>
</tr>
</thead>
<tbody>
<tr>
<td>Revenue: $50,000</td>
<td>Revenue: $65,000</td>
<td>Revenue: $72,000</td>
</tr>
<tr>
<td>Growth: 10%</td>
<td>Growth: 8%</td>
<td>Growth: 4%</td>
</tr>
</tbody>
</table>
count of "thead th" → 2, expected 3<table>
<caption>Quarterly revenue</caption>
<thead>
<tr scope="col">Q1</scope>
<th scope="col">Q2</th>
<th scope="col">Q3</th>
</tr>
</thead>
<tbody>
<tr>
<td>Revenue: $50,000</td>
<td>Revenue: $65,000</td>
<td>Revenue: $72,000</td>
</tr>
<tr>
<td>Growth: 10%</td>
<td>Growth: 8%</td>
<td>Growth: 4%</td>
</tr>
</tbody>
</table>
count of "thead th" → 2, expected 3<table>
<caption>Quarterly revenue</caption>
<thead>
<tr scope="col">Q1</scope>
<th scope="col">Q2</th>
<th scope="col">Q3</th>
</tr>
</thead>
<tbody>
<tr>
<td>Revenue: $50,000</td>
<td>Revenue: $65,000</td>
<td>Revenue: $72,000</td>
</tr>
<tr>
<td>Growth: 10%</td>
<td>Growth: 8%</td>
<td>Growth: 4%</td>
</tr>
</tbody>
</table>
balanced_depth — Max bracket nesting depth, -1 if unbalanced · 4 exampleshit the token cap mid-answer — compile error: task.cpp:8:1: error: 'We' does not name a type
8 | We are to compute the maximum nesting depth of round brackets in the string s.
| ^~
/tmp/bench_cpp_gqp79We are to compute the maximum nesting depth of round brackets in the string s.
Steps:
1. We traverse the string and ignore non-bracket characters.
2. We maintain a counter for the current depth (starting at 0).
3. When we see an opening bracket '(', we increase the counter by 1.
4. When we see a closing bracket ')', we decrease the counter by 1.
5. However, if at any point the counter becomes negative, that means there's a closing bracket without an opener -> not balanced -> return -1.
6. Also, at the end, if the counter is not zero, then there are unclosed openers -> not balanced -> return -1.
But note: the problem says "return the maximum nesting depth" if balanced, or -1 if not balanced.
However, note: the problem says "maximum nesting depth". We are to track the maximum value of the counter during the traversal.
Important: The problem states that non-bracket characters are ignored.
Approach:
Initialize:
current_depth = 0
max_depth = 0
balance = 0
For each char c in s:
if c is '(' -> balance++
then max_depth = max(max_depth, balance)
if c is ')' -> balance--
if balance < 0 -> return -1 (unbalanced)
else: ignore
After the loop, if balance != 0 -> return -1 (unclosed openers)
But note: the problem says the empty string has depth 0.
However, the problem says: "return the maximum nesting depth of round brackets in s"
Example:
s = "()" -> max_depth = 1?
Actually:
start: balance=0, max_depth=0
'(' -> balance=1, max_depth=1
')' -> balance=0 -> then max_de
… (truncated)hit the token cap mid-answer — no code in responseWe are to compute the maximum nesting depth of round brackets in the string s.
Steps:
1. We traverse the string and ignore non-bracket characters.
2. We maintain a counter for the current depth (starting at 0).
3. When we see an opening bracket '(', we increase the counter by 1.
4. When we see a closing bracket ')', we decrease the counter by 1.
5. However, if at any point the counter becomes negative, that means there's a closing bracket without an opener -> not balanced -> return -1.
6. Also, at the end, if the counter is not zero, then there are unclosed openers -> not balanced -> rehit the token cap mid-answer — no code in responseWe are to compute the maximum nesting depth of round brackets in the string s.
Steps:
1. We traverse the string.
2. We ignore non-bracket characters (only '(' and ')').
3. We maintain a counter for the current depth (starting at 0).
4. When we see an opening bracket '(', we increase the counter by 1 (and update the max depth if needed).
5. When we see a closing bracket ')', we decrease the counter by 1. But note: if the counter becomes negative, that means there's a closing bracket without an opener -> unbalanced -> return -1.
6. However, the problem says: return -1 if the brackets are hit the token cap mid-answer — no code in responseWe are to compute the maximum nesting depth of round brackets in the string s.
Steps:
1. We traverse the string and ignore non-bracket characters.
2. We maintain a counter for the current depth (starting at 0).
3. When we see an opening bracket '(', we increase the counter by 1.
4. When we see a closing bracket ')', we decrease the counter by 1.
5. However, if at any point the counter becomes negative, that means there's a closing bracket without an opener -> not balanced -> return -1.
6. Also, at the end, if the counter is not zero, then there are unclosed openers -> not balanced -> rehtml_nav_current — Nav list with aria-current on the active link · 4 exampleshit the token cap mid-answer — count of "nav" → 4, expected 1We are to create an HTML fragment without a document shell.
Requirements:
- A <nav> element
- Inside the nav, a <ul> of exactly three <li> items
- Each <li> holds an <a> tag
- The links point to: '/', '/about', '/contact'
- The About link (which is the link for '/about') must carry the attribute `aria-current="page"` and only it.
Steps:
- We'll create a <nav> element.
- Inside, a <ul> (unordered list).
- Three <li> elements, each with an <a> tag.
- The links:
first: href="/"
second: href="/about" -> this one should have aria-current="page"
third: href="/contact"
Important: Only the About link has `aria-current="page"`.
Note: The problem says "the About link — and only it — carries aria-current="page""
We'll structure the list items in the order:
- Home (href="/")
- About (href="/about", with aria-current="page")
- Contact (href="/contact")
However, note: the problem does not specify the order of the links, but typically we have Home, About, Contact.
But the problem says: "each holding an <a>". We have to have exactly three.
Let's write the fragment:
<nav>
<ul>
<li><a href="/">Home</a></li>
<li><a href="/about" aria-current="page">About</a></</li>
<li><a href="/contact">Contact</a></li>
</ul>
</nav>
But note: the problem says "the About link — and only it — carries aria-current="page"". So we set that attribute only on the About link.
However, the problem does not specify the text of the links. We can use the text "Home", "About", "Contact" for clarity.
But
… (truncated)hit the token cap mid-answer — count of "nav" → 6, expected 1We are to create an HTML fragment without a document shell.
Requirements:
- A <nav> element
- Inside the nav, a <ul> of exactly three <li> items
- Each <li> holds an <a> tag
- The links point to: '/', '/about', '/contact'
- The About link (which is the link for '/about') must carry the attribute `aria-current="page"` and only it.
Steps:
- We'll create a <nav> element.
- Inside, a <ul> (unordered list).
- Three <li> elements, each with an <a> tag.
- The links:
first: href="/"
second: href="/about" -> this one should have aria-current="page"
third: href="/contact"
Important: Only the About link has `aria-current="page"`.
Note: The problem says "the About link — and only it — carries aria-current="page""
We'll structure the list items in the order:
- Home (href="/")
- About (href="/about", with aria-current="page")
- Contact (href="/contact")
However, note: the problem does not specify the order of the links, but typically we have Home, About, Contact.
But the problem says: "each holding an <a>". We have to have exactly three.
Let's write the fragment:
<nav>
<ul>
<li><a href="/">Home</a></li>
<li><a href="/about" aria-current="page">About</a></</li>
<li><a href="/contact">Contact</a></li>
</ul>
</nav>
But note: the problem says "the About link — and only it — carries aria-current="page"". So we are only setting that on the About link.
However, the problem does not specify the text of the links. We can use the text "Home", "About", "Contact" for clarity.
But t
… (truncated)hit the token cap mid-answer — count of "nav" → 6, expected 1We are to create an HTML fragment without a document shell.
Requirements:
- A <nav> element
- Inside the nav, a <ul> of exactly three <li> items
- Each <li> holds an <a> tag
- The links point to: '/', '/about', '/contact'
- The About link (which is the link for '/about') must carry the attribute `aria-current="page"` and only it.
Steps:
- We'll create a <nav> element.
- Inside, a <ul> (unordered list).
- Three <li> elements, each with an <a> tag.
- The links:
first: href="/"
second: href="/about" -> this one should have aria-current="page"
third: href="/contact"
Important: Only the About link has `aria-current="page"`.
Note: The problem says "the About link — and only it — carries aria-current="page""
We'll structure the list items in the order:
- Home (href="/")
- About (href="/about", with aria-current="page")
- Contact (href="/contact")
However, note: the problem does not specify the order of the links, but typically we have Home, About, Contact.
But the problem says: "each holding an <a>". We have to have exactly three.
Let's write the fragment:
<nav>
<ul>
<li><a href="/">Home</a></li>
<li><a href="/about" aria-current="page">About</a></</li>
<li><a href="/contact">Contact</a></li>
</ul>
</nav>
But note: the problem says "the About link — and only it — carries aria-current="page"". So we are only setting that on the About link.
However, the problem does not specify the text of the links. We can use the text as "Home", "About", "Contact" for clarity?
Bu
… (truncated)hit the token cap mid-answer — count of "nav" → 5, expected 1We are to create an HTML fragment without a document shell.
Requirements:
- A <nav> element
- Inside the nav, a <ul> of exactly three <li> items
- Each <li> holds an <a> tag
- The links point to: '/', '/about', '/contact'
- The About link (which is the link for '/about') must carry aria-current="page"
Steps:
- We'll create the <nav> element.
- Inside, we'll have a <ul> (unordered list).
- We'll have three <li> elements, each with an <a> tag.
- The links:
first: href="/"
second: href="/about" -> this one should have aria-current="page"
third: href="/contact"
Note: The problem says "the About link — and only it — carries aria-current='page'".
So we set the aria-current="page" on the <a> for the about link.
Important: We are to return only the fragment in a single HTML code block.
Let's write the code accordingly.
Example structure:
<nav>
<ul>
<li><a href="/">Home</a></li>
<li><a href="/about" aria-current="page">About</a></</li>
<li><a href="/contact">Contact</a></li>
</ul>
</nav>
However, note: the problem does not specify the text of the links. But the problem says "the links point to ...", so we have to have the text?
But the problem says: "each holding an <a>". It doesn't specify the text, but typically we have text. However, the problem does not require text?
Let me re-read: "each holding an <a>". It doesn't say what the text is. But in the context of a navigation, we usually have text.
But the problem says: "Return only the fragment". And it doesn't s
… (truncated)sql_top_spenders — GROUP BY total with HAVING — not a WHERE · 4 exampleshit the token cap mid-answer — query returned "OperationalError: near \"We\": syntax error", expected [["cy", 150], ["ada", 110]]We are given a table `orders` with columns: customer (TEXT) and amount (INTEGER)
We want to return each customer whose total order amount (sum of all their amounts) is at least 100.
We need to output two columns: customer and total (the total amount for that customer)
The result should be ordered by total descending and then by customer ascending.
Steps:
1. We can use a subquery to compute the total amount per customer.
2. Then filter for total >= 100.
3. Order by total descending and customer ascending.
However, note: the problem says "return each customer whose ORDER TOTAL (sum of all their amounts) is at least 10:00" -> but the threshold is 100.
Important: The problem says "at least 100", so we use >= 100.
Let's write the query:
We'll use:
SELECT customer, total
FROM (
SELECT customer, SUM(amount) AS total
FROM orders
GROUP BY customer
) AS subquery
WHERE total >= 100
ORDER BY total DESC, customer ASC;
But note: the problem says "Return only the SQL in a single sql code block."
However, the problem says: "as two columns (customer, total)".
Let me check: the table has `customer` and `amount`. We are grouping by customer and summing the amount.
But note: the problem says "ORDER TOTAL (sum of all their amounts)".
We have to be cautious: the problem says "threshold applies to the customer's total, not to individual orders".
So the above query should work.
However, the problem says: "Write a single SQLite SELECT"
We can also do without a subquery by using a WHERE clause in the GROUP BY? But we need to filt
… (truncated)hit the token cap mid-answer — query returned "OperationalError: near \"We\": syntax error", expected [["cy", 150], ["ada", 110]]We are given a table `orders` with columns: customer (TEXT) and amount (INTEGER)
We want to return each customer whose total order amount (sum of all their amounts) is at least 100.
We need to output two columns: customer and total (the total amount for that customer)
The result should be ordered by total descending and then by customer ascending.
Steps:
1. We can use a subquery to compute the total amount per customer.
2. Then filter for total >= 100.
3. Order by total descending and customer ascending.
However, note: the problem says "return each customer whose ORDER TOTAL (sum of all their amounts) is at least 10:00" -> but the threshold is 100.
Important: The problem says "at least 100", so we use >= 100.
Let's write the query:
We'll use:
SELECT customer, total
FROM (
SELECT customer, SUM(amount) AS total
FROM orders
GROUP BY customer
) AS subquery
WHERE total >= 100
ORDER BY total DESC, customer ASC;
But note: the problem says "Return only the SQL in a single sql code block."
However, the problem says: "as two columns (customer, total)".
Let me check: the table has `customer` and `amount`. We are grouping by customer and summing the amount.
But note: the problem says "ORDER TOTAL (sum of all their amounts)".
We have to be cautious: the problem says "threshold applies to the customer's total, not to individual orders".
So the above query should work.
However, the problem says: "Write a single SQLite SELECT"
We can also do without a subquery by using a WHERE clause in the GROUP BY? But we need to filt
… (truncated)query returned "OperationalError: no such column: total", expected [["cy", 150], ["ada", 110]]SELECT customer, total FROM orders WHERE amount >= 100 ORDER BY total DESC, customer ASC;
query returned "OperationalError: no such column: total", expected [["cy", 150], ["ada", 110]]SELECT customer, total FROM orders WHERE amount >= 100 ORDER BY total DESC, customer ASC;
card_grid — Responsive auto-fill card grid · 4 examplesdecl → null, expected "translateY(-2px)".cards {
display: grid;
grid-template-columns: repeat(auto-fill, minmax(240px, 1fr));
gap: 16px;
}
.card {
border-radius: 8px;
transition: transform 0.3s;
cursor: pointer;
}
decl → null, expected "translateY(-2px)".cards {
display: grid;
grid-template-columns: repeat(auto-fill, minmax(240px, 1fr));
gap: 16px;
}
.card {
border-radius: 8px;
transition: transform 0.3s;
}
decl → null, expected "translateY(-2px)".cards {
display: grid;
grid-template-columns: repeat(auto-fill, minmax(240px, 1fr));
gap: 16px;
}
.card {
border-radius: 8px;
transition: transform 0.3s;
cursor: pointer;
}
decl → null, expected "translateY(-2px)".cards {
display: grid;
grid-template-columns: repeat(auto-fill, minmax(240px, 1fr));
gap: 16px;
}
.card {
border-radius: 8px;
transition: transform 0.3s;
}
GPU
NVIDIA GB10 · 122 GB
Unified memory
122 GB
CPU
Cortex-X925
Cores
20
OS
Ubuntu 24.04.4 LTS (aarch64)
Kernel
6.17.0-1018-nvidia
Ollama
0.32.5
Proxy
0.2.0
Every task asks for one answer in the task's language — Python, JavaScript under node, C and C++ under gcc, Rust, C#, or PHP, each run in a separate process under a timeout with the return value compared against the expected one. HTML and CSS tasks are graded structurally: the answer is parsed and checked against required structure (bindings, attributes, declarations in the right context) — a claim about the markup, not about how a browser renders it. Tasks whose toolchain is absent on the machine are skipped and listed here, never scored as zero. Fully correct counts only responses where every case for that task passed; cases is the share of individual cases that passed, so a near-miss still scores there. A response with no extractable code block scores zero — that measures instruction-following, not coding.
Suite
coding-v3
Tasks
47
Cases
298
Repeats
2 per task
Languages
12
Core — 15 tasks, 97 cases
group_ranges Collapse consecutive integers into range strings · jsclamp_add Saturating int addition — overflow trap · ccount_words Count words split on spaces and tabs · ccsv_escape RFC 4180 CSV field quoting · cppsnake_to_camel snake_case to camelCase — digits stop capitalisation · rustclamp_mul Saturating int multiplication · csharpslugify URL slug: symbol runs become one hyphen · phplogin_form Login form with labels bound to their inputs · htmlcard_grid Responsive auto-fill card grid · csssemver_cmp Semantic versions incl. pre-release precedencecsv_line Split one CSV record honouring quoteslru_ops LRU cache with eviction orderpath_norm Normalise a POSIX path with . and ..base_convert Integer between bases 2-36 with validationinterval_intersect Intersect two interval listsHard — 32 tasks, 201 cases
parse_query Parse a URL query string into an object · jsround_to Round to nearest multiple, halves away from zero · cbalanced_depth Max bracket nesting depth, -1 if unbalanced · cppmid_floor Floor midpoint of two i64s — overflow and negatives · rustordinal English ordinal suffix — the 11th/12th/13th trap · csharppluck Column from associative rows — null vs missing key · phpdata_table Revenue table with caption and scoped headers · htmltheme_vars Dark-mode token inside a media query · cssglob_match Glob matching with ? and *roman_strict Roman to int, rejecting non-canonical formstopo_lex Smallest topological order, None on cyclejustify Full text justificationjson_pointer Resolve an RFC 6901 JSON Pointertokenize_expr Tokenise arithmetic, None on invalid inputquoted_split Shell-style split honouring quotes and escapesring_buffer Ring buffer that overwrites the oldest on overflowexpand_ranges Expand "1-3,7" specs with validationdedent_text Longest common whitespace prefix — tabs are not spacesdeep_get Path lookup like a.b[2].c, null on any miss · jssql_top_spenders GROUP BY total with HAVING — not a WHERE · sqlsql_missing_users Anti-join where NOT IN meets a NULL · sqlbash_dedup_lines First occurrence wins, order preserved (sort -u reorders) · bashbash_sort_versions Numeric per-component version sort · bashgo_rle_decode Decode letter+count runs, strict validation · gogo_ipv4_valid Strict dotted-quad IPv4 — leading zeros invalid · goc_bit_count_range Popcount over [a,b] — i <= INT_MAX loop overflow trap · ccpp_wrap_count Greedy word-wrap line count — long words never split · cpprust_kv_get k=v parsing: first '=' splits, last duplicate wins · rustcs_round_half Half away from zero — Math.Round is banker's · csharpphp_flatten_keys Nested assoc array to dot-notation keys · phphtml_nav_current Nav list with aria-current on the active link · htmlcss_sticky_footer Flexbox sticky footer, no fixed positioning · cssx-client-name: ai-proxy-bench.Time for the discarded warm-up request — the price of making the model resident, excluded from every measurement above.
| Configuration | Cold start |
|---|---|
| devstral-2:123b · ollama · cached · 4 | 265.3 s |
| devstral-2:123b · ollama · cached · 1 | 264.6 s |
| llama3:70b-instruct · ollama · cached · 1 | 136.3 s |
| llama3:70b-instruct · ollama · cached · 4 | 133.2 s |
| gemma3:27b · ollama · cached · 1 | 90.9 s |
| gemma3:27b · ollama · cached · 4 | 90.8 s |
| qwen3.6:27b · ollama · cached · 1 | 86.1 s |
| qwen3.6:27b · ollama · cached · 4 | 86.1 s |
| codellama:70b · ollama · cached · 1 | 85.7 s |
| devstral-small-2:24b · ollama · cached · 1 | 73.1 s |
+ 86 more under 73.1 s.
Screen-only companion to the annotated chart above: hover a dot for its numbers, drag a box to zoom into the crowded band, double-click to reset. Hovering a model anywhere on this page highlights it everywhere. The printed report keeps the annotated version.