Across e/accCounts toward ranking
Evidence budget4%
Upstream data time10/16 02:52
Last successful sync10/17 08:05
What it measures and how
Use only mean_score from Epoch’s own reruns at a fixed reasoning tier. Do not read ECI or aggregate Best score columns.
How this evidence is used
Counts toward ranking; benchmarks in the same task family share a fixed budget.
Benchmark results
Fixed rules select one representative configuration per public model. Anonymous test models are omitted. Original source ranks are retained, with up to 30 models per benchmark.
| Source rank | Source model name | Raw result | Representative configuration |
|---|---|---|---|
| 1 | gemini-4-argon_highGoogle | 76.0% | High reasoning |
| 2 | gpt-6-astra_maxOpenAI | 75.6% | Max reasoning |
| 3 | gpt-6.1-sol_maxOpenAI | 73.9% | Max reasoning |
| 4 | gemini-3.1-pro-preview_highGoogle | 73.5% | High reasoning |
| 5 | claude-opus-5-5_maxAnthropic | 72.2% | Max reasoning |
| 6 | claude-fable-5-1_maxAnthropic | 70.8% | Max reasoning |
| 7 | claude-fable-5_xhighAnthropic | 70.7% | xHigh reasoning |
| 8 | gpt-5.6-sol_maxOpenAI | 69.7% | Max reasoning |
| 8 | gemini-3.8-flash_highGoogle | 69.7% | High reasoning |
| 10 | gemini-3.7-flash_highGoogle | 69.2% | High reasoning |
| 11 | gemini-3-flash-preview_highGoogle | 66.8% | High reasoning |
| 12 | gemini-3.6-flash_highGoogle | 66.2% | High reasoning |
| 12 | gemini-3.5-flash_highGoogle | 66.2% | High reasoning |
| 14 | gpt-5.5_xhighOpenAI | 63.0% | xHigh reasoning |
| 15 | gpt-6-sol_maxOpenAI | 60.7% | Max reasoning |
| 16 | muse-spark-1.2_xhighMeta | 60.3% | xHigh reasoning |
| 17 | claude-opus-5_maxAnthropic | 59.9% | Max reasoning |
| 18 | muse-spark-1.1Meta | 57.8% | Source default configuration |
| 19 | Muse-Spark-Open-816B-A42B_xhighMeta AI | 56.6% | xHigh reasoning |
| 20 | grok-4.7_xhighxAI | 56.0% | xHigh reasoning |
| 21 | qwen3.7-maxAlibaba | 55.8% | Source default configuration |
| 22 | claude-opus-4-8_maxAnthropic | 53.0% | Max reasoning |
| 23 | deepseek-v4-pro-0813_maxDeepSeek | 52.9% | Max reasoning |
| 24 | qwen3.6-max-previewAlibaba | 52.0% | Source default configuration |
| 25 | claude-opus-4-7_xhighAnthropic | 51.7% | xHigh reasoning |
| 26 | kimi-k3_maxMoonshot AI | 50.6% | Max reasoning |
| 27 | gpt-5-2025-08-07_highOpenAI | 50.1% | High reasoning |
| 28 | o3-2025-04-16_highOpenAI | 49.4% | High reasoning |
| 30 | grok-4.6_xhighxAI | 48.9% | xHigh reasoning |
| 31 | qwen3-max-2025-09-23Alibaba | 48.7% | Source default configuration |
Limitations and attribution
Counts toward ranking; benchmarks in the same task family share a fixed budget.
Data license: CC BY 4.0 · Data evaluated by Epoch AI itself
Results are published by Epoch AI. Raw scores and the e/acc consensus index use different scales and cannot be added directly.