Frontiermath tiers 1 3 v2 — Epoch AI
Frontiermath tiers 1 3 v2 — Epoch AI is run by Epoch AI. It has ranked 106 models, of which the catalogue holds 72, scoring from 0 to 93.684.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 1stof 106 | GPT 6 Astra | mean_score (max) | 93.684 |
| 2ndof 106 | Claude Fable 5.1 | mean_score (max) | 90.175 |
| 3rdof 106 | GPT-5.6 Sol | mean_score (max) | 89.123 |
| 4thof 106 | GPT-5.5 Pro | mean_score (xhigh) | 87.719 |
| 5thof 106 | Claude Fable 5 | mean_score (max) | 87.018 |
| 6thof 106 | GPT-5.6 Terra | mean_score (max) | 85.965 |
| 7thof 106 | Claude Opus 5 | mean_score (max) | 85.614 |
| 8thof 106 | GPT-5.5 | mean_score (xhigh) | 85.263 |
| 9thof 106 | GPT-5.4 Pro | mean_score (xhigh) | 82.456 |
| 10thof 106 | GPT-5.6 Luna | mean_score (max) | 82.105 |
| 11thof 106 | Claude Opus 4.8 | mean_score (max) | 80 |
| 12thof 106 | GPT-5.4 | mean_score (xhigh) | 78.596 |
| 13thof 106 | Qwen 3.8 Max | mean_score (xhigh) | 74.737 |
| 14thof 106 | GPT-5.2 Pro | mean_score (xhigh) | 74 |
| 15thof 106 | Kimi K3 | mean_score (max) | 72.183 |
| 16thof 106 | Gemini 3.7 Flash | mean_score (high) | 71.579 |
| 17thof 106 | Claude Opus 4.7 | mean_score (max) | 70.175 |
| 18thof 106 | GLM-5.3 | mean_score (max) | 68.772 |
| 19thof 106 | Gemini 3.8 Flash | mean_score (high) | 68.421 |
| 20thof 106 | GPT-5.2 | mean_score (xhigh) | 67.4 |
| 21stof 106 | Grok 4.6 | mean_score (xhigh) | 65.965 |
| 22ndof 106 | Claude Opus 4.6 | mean_score (max) | 65.965 |
| 24thof 106 | Claude Sonnet 5 | mean_score (max) | 65.614 |
| 25thof 106 | DeepSeek V4 Pro 0813 | mean_score (max) | 64.561 |
| 26thof 106 | Qwen3.7 Max | mean_score (max) | 64.561 |
| 27thof 106 | Gemini 3.5 Flash | mean_score (high) | 62.807 |
| 28thof 106 | Gemini 3.1 Pro Preview | mean_score | 59.649 |
| 29thof 106 | GLM 5.2 | mean_score (max) | 59.206 |
| 30thof 106 | Gemini 3.6 Flash | mean_score (high) | 58.947 |
| 31stof 106 | DeepSeek V4 Flash (0731) | mean_score (max) | 57.544 |
| 32ndof 106 | Grok 4.5 | mean_score (high) | 57.193 |
| 33rdof 106 | Kimi K2.6 | mean_score | 57.193 |
| 34thof 106 | GLM 5.3 Flash | mean_score (max) | 55.789 |
| 35thof 106 | GPT-5 Pro | mean_score (high) | 55.789 |
| 36thof 106 | GPT-5 | mean_score (high) | 55.439 |
| 38thof 106 | Kimi K2.7 Code | mean_score | 54.035 |
| 39thof 106 | GPT-5.4 mini | mean_score (xhigh) | 51.228 |
| 40thof 106 | Gemini 3 Flash Preview | mean_score | 51.228 |
| 41stof 106 | GPT-5 mini | mean_score (high) | 46.667 |
| 42ndof 106 | Inkling Small | mean_score (xhigh) | 46.316 |
| 43rdof 106 | DeepSeek V4 Pro | mean_score (max) | 45.263 |
| 44thof 106 | Grok 4.20 | mean_score | 44.912 |
| 45thof 106 | GPT-5.4 nano | mean_score (high) | 44.912 |
| 46thof 106 | Grok 4.3 | mean_score (high) | 42.807 |
| 50thof 106 | Qwen3.6 Plus | mean_score | 38.246 |
| 52ndof 106 | GLM 5.1 | mean_score | 36.842 |
| 53rdof 106 | o4-mini | mean_score (high) | 36.14 |
| 54thof 106 | Qwen3.6 27B | mean_score | 35.088 |
| 55thof 106 | Qwen 3.7 Plus | mean_score (none) | 34.386 |
| 56thof 106 | Claude Opus 4.5 | mean_score | 34.386 |
| 58thof 106 | o3 | mean_score (high) | 33.333 |
| 59thof 106 | Inkling | mean_score (xhigh) | 33.333 |
| 61stof 106 | Qwen3.5 397B A17B | mean_score (none) | 31.228 |
| 65thof 106 | Gemini 3.1 Flash-Lite | mean_score (high) | 27.719 |
| 66thof 106 | GPT-5.5 Instant | mean_score | 26.316 |
| 67thof 106 | Gemini 3.5 Flash-Lite | mean_score (high) | 25.965 |
| 70thof 106 | Gemini 2.5 Pro | mean_score | 24.561 |
| 71stof 106 | Claude Sonnet 4.5 | mean_score | 23.86 |
| 73rdof 106 | Qwen3.6-Flash | mean_score (none) | 22.456 |
| 75thof 106 | Qwen3.6 35B A3B | mean_score (none) | 20.351 |
| 77thof 106 | GPT-5 nano | mean_score (high) | 20 |
| 78thof 106 | Qwen3.7 Flash | mean_score (none) | 19.298 |
| 82ndof 106 | o3-mini | mean_score (high) | 18.596 |
| 83rdof 106 | Qwen3.5-Flash | mean_score (none) | 18.246 |
| 90thof 106 | o1 | mean_score (high) | 14.737 |
| 91stof 106 | Claude Opus 4.1 | mean_score | 12.632 |
| 96thof 106 | GPT-4.1 mini | mean_score | 6.667 |
| 97thof 106 | GPT-4.1 fine-tuned | mean_score | 5.965 |
| 103rdof 106 | GPT-4o-mini (2024-07-18) | mean_score | 70.2 |
| 104thof 106 | GPT-4 Turbo | mean_score | 70.2 |
| 105thof 106 | GPT-4o (2024-08-06) | mean_score | 35.1 |
| 106thof 106 | GPT-3.5 Turbo | mean_score | 0 |
Read from the board on 2026-09-16