CAE-Bench RC7 - Varying Physical Inputs: leaderboard
Metric: Score (0-100), 56 tasks. Source: arxiv.org. Saturation forecast: Around July 2027. 12 models tracked. Updated 18 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-6.1 Sol | Medium | 50.35 | #12 |
| 2 | GPT-6 Astra | Medium | 49.79 | #7 |
| 3 | Gemini 4 Argon | Medium | 48.2 | #14 |
| 4 | GPT-5.6 Sol | Medium | 44.37 | #33 |
| 5 | GPT-5.6 Luna | Max | 43.56 | #129 |
| 6 | GPT-6 Luna | xHigh | 42.25 | #102 |
| 7 | Gemini 3.8 Flash | Medium | 41.06 | #39 |
| 8 | GPT-6 Sol | Medium | 39.9 | #30 |
| 9 | DeepSeek V4.1 Flash | High | 32.73 | #67 |
| 10 | Qwen 3.8 Flash | Low | 30.16 | #127 |
| 11 | Kimi K2.7 Code | Thinking | 24.85 | #195 |
| 12 | GLM-5.3 Flash | Low | 21.38 | #134 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=cae-bench-rc7-varying-physical-inputs · How It Works · Data refreshed daily, snapshot 2026-11-18.