Benchmarks
Epoch Capabilities Index
Models scored
462
Best score
161.7
Metric
ECI Score
Epoch Capabilities Index is an AI evaluation tracked by Epoch AI, with 462 scored model versions on record. Epoch AI runs this evaluation directly, so the scores are reproducible against their published logs. It is not a single test but an index: it combines results across many benchmarks into one capability score, which keeps rising even as individual benchmarks saturate.
Scored models span February 2023 to July 2026. The best score recorded is 161.72, measured as ECI Score. Models come from Anthropic, OpenAI, Moonshot, DeepSeek, Google, xAI, Z.ai (Zhipu AI), Alibaba and others.
Full record
- Score metric
- ECI Score
- Models scored
- 462
- Best score recorded
- 161.7
- Earliest model scored
- Feb 24, 2023
- Most recent model scored
- Jul 24, 2026
- Third-party reported
- No
Leaderboard
| 01 | GPT-5.6 Sol (medium) | OpenAI | 161.7 |
| 02 | GPT-5.5 Pro (xhigh) | OpenAI | 160.9 |
| 03 | Claude Fable 5 (max) | Anthropic | 160.7 |
| 04 | Claude Opus 5 | Anthropic | 159.4 |
| 05 | GPT-5.5 (medium) | OpenAI | 158.5 |
| 06 | GPT-5.6 Terra (low) | OpenAI | 158.2 |
| 07 | Claude Opus 4.8 | Anthropic | 158 |
| 08 | GPT-5.4 Pro (no thinking) | OpenAI | 157.9 |
| 09 | GPT-5.4 (high) | OpenAI | 156.1 |
| 10 | Claude Opus 4.7 (max) | Anthropic | 156.1 |
| 11 | GPT-5.3 Codex (xhigh) | OpenAI | 155.6 |
More benchmarks
SourceEpoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/benchmarks. Licensed under CC BY 4.0.
← All benchmarks