Benchmarks · reasoning
CritPt
CritPt: research-level physics reasoning problems; accuracy.
138 models · 269 results · data as of 2026-10-29 · artificialanalysis.ai ↗
Leaderboard
Gemini 4 Carbon high
More from this benchmark
3 published measuresScore vs any published measure3 measures: sub-scores, splits, counts
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Gemini 4 Carbon high | 21,942,908 | 415,698 | 11,510,951 |
| GPT-5.6 Sol max | 257,255 | 128,393 | 2,827,158 |
| Opus 5.5 max, fallback | 342,494 | 211,533 | 7,383,628 |
| GPT-6 Astra max | 258,783 | 140,415 | 1,440,455 |
| GPT-6.1 Sol max | 267,530 | 148,981 | 1,494,413 |
| Sonnet 5.5 max, fallback | 348,175 | 223,585 | 12,947,380 |
| Fable 5.1 xhigh, fallback | 443,566 | 279,949 | 5,978,590 |
| GPT-6 Sol max | 206,320 | 86,622 | 2,142,738 |
| GPT-5.5 Pro xhigh | 449,002 | 93,054 | 2,205,227 |
| GPT-5.4 Pro xhigh | 139,482 | 241,835 | 5,636,744 |
| GPT-5.6 Terra max | 1,080,638 | 106,726 | 3,262,212 |
| Opus 5 max | 385,422 | 254,953 | 7,744,686 |
| GPT-6.1 Terra max | 231,015 | 106,219 | 3,078,294 |
| Gemini 4 Argon high | 17,098,697 | 294,036 | 7,373,276 |
| GPT-5.5 xhigh | 139,482 | 162,341 | 3,579,015 |
| MiMo V2.6 Pro | 299,257 | 457,989 | 4,252,199 |
| Muse Spark 1.3 xhigh | 286,040 | 173,103 | 4,815,850 |
| DeepSeek V4.1 Flash 1017 max | 257,438 | 131,903 | 9,060,456 |
| GPT-5.4 xhigh | 139,482 | 237,621 | 5,446,950 |
| Kimi K3 max | 309,859 | 103,871 | 4,042,794 |
| Haiku 5.5 xhigh | 354,606 | 208,497 | 8,836,041 |
| Opus 4.8 max | 363,585 | 188,267 | 5,163,431 |
| GLM 5.2 max | 708,844 | 105,770 | 7,343,085 |
| GPT-5.6 Luna max | 1,416,201 | 95,607 | 3,362,194 |
| Qwen 3.8 2.4T-A95B | 346,085 | 204,254 | 5,312,960 |
All results
269 rows| # | Model | Effort | CritPt |
|---|---|---|---|
| 1 | Gemini 4 Carbon | high | 36.0% |
| 2 | GPT-5.6 Sol | max | 32.3% |
| 3 | Opus 5.5 | max | 31.7% |
| 4 | Opus 5.5 | xhigh | 31.7% |
| 5 | GPT-6 Astra | max | 31.7% |
| 6 | GPT-6.1 Sol | max | 31.7% |
| 7 | GPT-6.1 Sol | xhigh | 31.7% |
| 8 | Sonnet 5.5 | max | 31.4% |
| 9 | GPT-6 Astra | xhigh | 31.4% |
| 10 | Fable 5.1 | xhigh | 31.1% |
| 11 | Sonnet 5.5 | xhigh | 31.1% |
| 12 | Opus 5.5 | high | 30.9% |
| 13 | GPT-6 Sol | max | 30.9% |
| 14 | GPT-5.5 Pro | xhigh | 30.6% |
| 15 | Fable 5.1 | high | 30.3% |
| 16 | GPT-5.4 Pro | xhigh | 30.0% |
| 17 | GPT-5.6 Terra | max | 30.0% |
| 18 | GPT-6.1 Sol | high | 30.0% |
| 19 | Fable 5.1 | max | 29.7% |
| 20 | Fable 5.1 | medium | 29.1% |
| 21 | Opus 5 | max | 29.1% |
| 22 | GPT-6 Astra | medium | 29.1% |
| 23 | GPT-6 Astra | high | 28.9% |
| 24 | GPT-5.6 Sol | xhigh | 28.6% |
| 25 | Opus 5 | high | 28.3% |
| 26 | GPT-6 Sol | xhigh | 28.0% |
| 27 | GPT-6.1 Terra | max | 28.0% |
| 28 | Fable 5.1 | low | 27.7% |
| 29 | Opus 5.5 | medium | 27.7% |
| 30 | Opus 5 | xhigh | 27.7% |