Arc agi 2 — Epoch AI
Arc agi 2 — Epoch AI is run by Epoch AI. It has ranked 290 models, of which the catalogue holds 88, scoring from 0 to 95.
measured by Epoch AI · the board itself
| Place | Model | Metric | Score |
|---|---|---|---|
| 1stof 290 | GPT 6 Astra | Score (max) | 95 |
| 2ndof 290 | GPT 6.1 Sol | Score (max) | 94.167 |
| 4thof 290 | Claude Opus 5.5 | Score (high) | 93.333 |
| 5thof 290 | GPT-5.6 Sol | Score (max) | 92.5 |
| 12thof 290 | Gemini 4 Argon | Score (high) | 91.25 |
| 13thof 290 | Claude Opus 5 | Score (max) | 90.42 |
| 16thof 290 | Claude Fable 5.1 | Score (max) | 90 |
| 18thof 290 | GPT 6 Sol | Score (max) | 89.583 |
| 19thof 290 | Claude Fable 5 | Score (max) | 89.167 |
| 20thof 290 | Gemini 3.8 Flash | Score (high) | 89.167 |
| 32ndof 290 | GPT-5.5 | Score (xhigh) | 85 |
| 34thof 290 | Gemini 3.7 Flash | Score (high) | 84.583 |
| 35thof 290 | Gemini 3 Deep Think | Score | 84.58 |
| 36thof 290 | GPT-5.5 Pro | Score (high) | 84.58 |
| 39thof 290 | GPT-5.6 Terra | Score (max) | 83.9 |
| 40thof 290 | Claude Sonnet 5.5 | Score (xhigh) | 83.889 |
| 41stof 290 | GPT-5.4 Pro | Score (xhigh) | 83.33 |
| 48thof 290 | DeepSeek V4.1 Flash 1017 | Score (max) | 78.056 |
| 50thof 290 | Gemini 3.1 Pro Preview | Score | 77.1 |
| 54thof 290 | Claude Opus 4.7 | Score (max) | 75.83 |
| 58thof 290 | GPT-5.4 | Score (xhigh) | 73.95 |
| 61stof 290 | DeepSeek V4.1 Flash | Score (max) | 72.917 |
| 62ndof 290 | Gemini 3.5 Flash | Score (high) | 72.08 |
| 63rdof 290 | Claude Opus 4.8 | Score (high) | 72.08 |
| 67thof 290 | Claude Opus 4.6 | Score | 69.17 |
| 77thof 290 | Grok 4.6 | Score (xhigh) | 67.083 |
| 81stof 290 | GLM 5.3 Flash | Score (max) | 65.833 |
| 82ndof 290 | Grok 4.20 | Score | 65.14 |
| 88thof 290 | DeepSeek V4 Flash (0731) | Score (max) | 61.389 |
| 89thof 290 | Grok 4.7 | Score (xhigh) | 61.389 |
| 91stof 290 | DeepSeek V4 Pro 0813 | Score (max) | 61.25 |
| 92ndof 290 | Claude Haiku 5.5 | Score (xhigh) | 61.25 |
| 94thof 290 | Claude Sonnet 4.6 | Score (high) | 60.42 |
| 95thof 290 | Kimi K3 | Score (max) | 60.417 |
| 96thof 290 | Gemini 3.6 Flash | Score (high) | 60.417 |
| 99thof 290 | GPT-5.6 Luna | Score (max) | 59.54 |
| 100thof 290 | GPT 6 Luna | Score (max) | 59.306 |
| 105thof 290 | Muse-Spark-Open-816B-A42B | Score (xhigh) | 57.5 |
| 110thof 290 | GPT-5.2 Pro | Score (high) | 54.16 |
| 112thof 290 | GPT-5.2 | Score (xhigh) | 52.91 |
| 115thof 290 | Grok 4.5 | Score (high) | 52.64 |
| 126thof 290 | Qwen3.8 27B | Score (xhigh) | 42.361 |
| 129thof 290 | Inkling Small | Score (xhigh) | 40.139 |
| 131stof 290 | Claude Opus 4.5 | Score | 37.64 |
| 133rdof 290 | Inkling | Score | 36.528 |
| 134thof 290 | Gemini 3 Flash Preview | Score | 33.61 |
| 141stof 290 | gemini-3-pro-preview | Score | 31.11 |
| 154thof 290 | GLM 5.2 | Score | 22.775 |
| 156thof 290 | GPT-5.4 mini | Score (xhigh) | 18.9 |
| 159thof 290 | GPT-5 Pro | Score | 18.33 |
| 161stof 290 | GPT-5.1 | Score (high) | 17.64 |
| 166thof 290 | Claude Sonnet 4.5 | Score | 13.61 |
| 172ndof 290 | Kimi K2.5 | Score | 11.81 |
| 173rdof 290 | Gemini 3.5 Flash-Lite | Score (high) | 10.278 |
| 174thof 290 | GPT-5 | Score (high) | 9.86 |
| 178thof 290 | Claude Opus 4 | Score | 8.61 |
| 184thof 290 | o3 | Score (high) | 6.53 |
| 188thof 290 | o4-mini | Score (high) | 6.11 |
| 189thof 290 | Claude Sonnet 4 | Score | 5.93 |
| 191stof 290 | GPT-5.4 nano | Score (xhigh) | 5.69 |
| 193rdof 290 | Grok 4 Fast | Score | 5.28 |
| 197thof 290 | o3-pro | Score (high) | 4.86 |
| 198thof 290 | Gemini 2.5 Pro | Score | 4.86 |
| 199thof 290 | MiniMax M2.5 | Score | 4.86 |
| 200thof 290 | GLM-5 | Score | 4.86 |
| 203rdof 290 | GPT-5 mini | Score (high) | 4.44 |
| 207thof 290 | Claude Haiku 4.5 | Score | 4.03 |
| 208thof 290 | DeepSeek v3.2 | Score | 4.03 |
| 213thof 290 | o3-mini | Score (high) | 2.99 |
| 220thof 290 | GPT-5 nano | Score (high) | 2.61 |
| 221stof 290 | Gemini 2.5 Flash | Score | 2.54 |
| 245thof 290 | gemini-2.0-flash-001 | Score | 1.3 |
| 246thof 290 | DeepSeek-R1 | Score | 1.3 |
| 248thof 290 | Codex mini | Score | 1.27 |
| 250thof 290 | Qwen3-235B-A22B-Instruct-2507 | Score | 1.25 |
| 255thof 290 | DeepSeek-R1 (0528) | Score | 1.12 |
| 257thof 290 | Claude 3.7 Sonnet | Score | 90 |
| 262ndof 290 | o1-mini | Score | 83 |
| 268thof 290 | GPT-4.1 fine-tuned | Score | 42 |
| 269thof 290 | Grok 3 Mini | Score (low) | 42 |
| 273rdof 290 | GPT-4o-mini (2024-07-18) | Score | 0 |
| 274thof 290 | GPT-4.1 nano | Score | 0 |
| 276thof 290 | Llama-4-Maverick-17B-128E-Instruct | Score | 0 |
| 278thof 290 | GPT-4.1 mini | Score | 0 |
| 279thof 290 | Grok 3 | Score | 0 |
| 281stof 290 | Magistral Small 2506 | Score | 0 |
| 283rdof 290 | Gemini 2.5 Pro Preview 06-05 | Score | 0 |
| 286thof 290 | GPT-4o (2024-11-20) | Score | 0 |
Read from the board on 2026-10-31