Apex agents — Epoch AI
Apex agents — Epoch AI is run by Epoch AI. It has ranked 65 models, of which the catalogue holds 52, scoring from 0.011 to 0.474.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 1stof 65 | Claude Fable 5.1 | Pass@1 score | 0.474 |
| 2ndof 65 | GPT 6 Astra | Pass@1 score | 0.467 |
| 3rdof 65 | Claude Fable 5 | Pass@1 score | 0.45 |
| 5thof 65 | Claude Opus 5 | Pass@1 score (max) | 0.435 |
| 6thof 65 | Claude Opus 4.8 | Pass@1 score (max) | 0.425 |
| 7thof 65 | Muse Spark 1.1 | Pass@1 score | 0.419 |
| 8thof 65 | Grok 4.6 | Pass@1 score | 0.412 |
| 10thof 65 | GPT-5.6 Sol | Pass@1 score (max) | 0.399 |
| 11thof 65 | Kimi K3 | Pass@1 score | 0.393 |
| 12thof 65 | GPT-5.5 | Pass@1 score (xhigh) | 0.385 |
| 16thof 65 | GPT-5.4 | Pass@1 score (xhigh) | 0.36 |
| 17thof 65 | GLM 5.2 | Pass@1 score | 0.356 |
| 19thof 65 | GPT-5.2 | Pass@1 score (xhigh) | 0.344 |
| 20thof 65 | Grok 4.5 | Pass@1 score | 0.342 |
| 21stof 65 | Claude Opus 4.7 | Pass@1 score (max) | 0.339 |
| 22ndof 65 | Gemini 3.1 Pro Preview | Pass@1 score | 0.335 |
| 23rdof 65 | Claude Sonnet 5 | Pass@1 score | 0.325 |
| 24thof 65 | Claude Opus 4.6 | Pass@1 score | 0.324 |
| 26thof 65 | GPT-5.3 Codex | Pass@1 score | 0.318 |
| 27thof 65 | gemini-3-pro-preview | Pass@1 score | 0.315 |
| 28thof 65 | GPT-5.2 Codex | Pass@1 score | 0.276 |
| 29thof 65 | Kimi K2.7 Code | Pass@1 score | 0.276 |
| 30thof 65 | GPT-5.4 mini | Pass@1 score | 0.246 |
| 31stof 65 | Gemini 3 Flash Preview | Pass@1 score | 0.24 |
| 32ndof 65 | Claude Sonnet 4.6 | Pass@1 score (high) | 0.237 |
| 35thof 65 | GPT-5.1 Codex | Pass@1 score | 0.207 |
| 36thof 65 | Claude Opus 4.5 | Pass@1 score | 0.207 |
| 37thof 65 | GPT-5 Codex (batch) | Pass@1 score (high) | 0.201 |
| 38thof 65 | Kimi K2.6 | Pass@1 score | 0.189 |
| 39thof 65 | GPT-5 | Pass@1 score (high) | 0.183 |
| 41stof 65 | GPT-5.1 | Pass@1 score (high) | 0.175 |
| 43rdof 65 | o3 | Pass@1 score (high) | 0.172 |
| 44thof 65 | GLM-5 | Pass@1 score | 0.172 |
| 45thof 65 | GPT-5.4 nano | Pass@1 score | 0.169 |
| 47thof 65 | Kimi K2.5 | Pass@1 score | 0.144 |
| 48thof 65 | Qwen 3.5 Plus | Pass@1 score | 0.136 |
| 49thof 65 | Gemini 3.1 Flash-Lite | Pass@1 score | 0.13 |
| 50thof 65 | Grok 4.1 | Pass@1 score | 0.128 |
| 51stof 65 | Nemotron 3 Ultra | Pass@1 score | 0.115 |
| 52ndof 65 | Claude Sonnet 4 | Pass@1 score | 0.093 |
| 53rdof 65 | Claude Haiku 4.5 | Pass@1 score | 0.089 |
| 54thof 65 | GLM-4.7 | Pass@1 score | 0.087 |
| 56thof 65 | DeepSeek v3.2 | Pass@1 score | 0.07 |
| 57thof 65 | Gemini 2.5 Pro | Pass@1 score | 0.066 |
| 58thof 65 | MiniMax M2.5 | Pass@1 score | 0.062 |
| 59thof 65 | GPT OSS 120B | Pass@1 score | 0.047 |
| 60thof 65 | Kimi K2 Thinking | Pass@1 score | 0.041 |
| 61stof 65 | GLM-4.6 | Pass@1 score | 0.04 |
| 62ndof 65 | Grok 3 | Pass@1 score | 0.021 |
| 63rdof 65 | Gemini 2.5 Flash | Pass@1 score | 0.018 |
| 64thof 65 | o1 | Pass@1 score (high) | 0.011 |
| 65thof 65 | GPT-4o (2024-11-20) | Pass@1 score | 0.011 |
Read from the board on 2026-09-16