Ebr bench — Epoch AI
Ebr bench — Epoch AI is run by Epoch AI. It has ranked 28 models, of which the catalogue holds 27, scoring from 2.381 to 78.095.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 1stof 28 | GPT-6.1 Terra | mean_score (max) | 78.095 |
| 2ndof 28 | GPT 6 Astra | mean_score (max) | 76.19 |
| 3rdof 28 | Claude Opus 5.5 | mean_score (max) | 71.429 |
| 4thof 28 | Gemini 4 Argon | mean_score (high) | 64.762 |
| 5thof 28 | Claude Fable 5.1 | mean_score (max) | 57.143 |
| 6thof 28 | GPT 6.1 Sol | mean_score (max) | 54.286 |
| 7thof 28 | GPT 6 Sol | mean_score (max) | 53.333 |
| 8thof 28 | Claude Opus 5 | mean_score (max) | 45.714 |
| 9thof 28 | GPT-5.6 Sol | mean_score (max) | 44.762 |
| 10thof 28 | Claude Haiku 5.5 | mean_score (max) | 43.81 |
| 11thof 28 | Claude Fable 5 | mean_score (max) | 39.524 |
| 12thof 28 | GPT-5.5 | mean_score (xhigh) | 34.286 |
| 14thof 28 | Grok 4.6 | mean_score (xhigh) | 30.476 |
| 15thof 28 | Claude Opus 4.8 | mean_score (max) | 28.571 |
| 16thof 28 | GPT-5.4 | mean_score (xhigh) | 25.397 |
| 17thof 28 | GPT-5.2 | mean_score (xhigh) | 23.016 |
| 18thof 28 | Claude Opus 4.7 | mean_score (max) | 19.048 |
| 19thof 28 | Claude Opus 4.5 | mean_score | 14.286 |
| 20thof 28 | Gemini 3.1 Pro Preview | mean_score | 14.286 |
| 21stof 28 | Claude Opus 4.6 | mean_score (max) | 12.698 |
| 22ndof 28 | GPT-5 | mean_score (high) | 12.698 |
| 23rdof 28 | GLM 5.2 | mean_score (max) | 9.524 |
| 24thof 28 | Qwen3.7 Max | mean_score (max) | 9.524 |
| 25thof 28 | Claude Opus 4.1 | mean_score | 7.937 |
| 26thof 28 | Gemini 3.5 Flash | mean_score (high) | 4.762 |
| 27thof 28 | Kimi K2.6 | mean_score | 2.381 |
| 28thof 28 | Claude Sonnet 4.5 | mean_score | 2.381 |
Read from the board on 2026-10-31