Benchmarks · reasoning
CritPt v2
Uses Artificial Analysis accuracy on the revised research-level physics benchmark, evaluated with code execution. Score: accuracy. Collected by Jake Boggs' benchmark dashboard.
11 models · 11 results · data as of 2026-10-29 · artificialanalysis.ai ↗
Leaderboard
GPT-6 Astra max
All results
11 rows| # | Model | Effort | CritPt v2 |
|---|---|---|---|
| 1 | GPT-6 Astra | max | 83.1% |
| 2 | GPT-6.1 Sol | max | 82.0% |
| 3 | Opus 5.5 | max | 77.7% |
| 4 | GPT-5.6 Sol | max | 74.9% |
| 5 | Muse Wonder Full | max | 73.1% |
| 6 | Gemini 4 Argon | high | 72.9% |
| 7 | Fable 5.1 | max | 72.0% |
| 8 | Muse Wonder (Max) | max | 69.7% |
| 9 | Grok 4.7 | xhigh | 50.6% |
| 10 | DeepSeek V4.1 Flash 1017 | max | 50.0% |
| 11 | GPT-5 | high | 28.9% |