Benchmarks / External Composite Indices
Epoch Capabilities Index
Epoch AI's composite capabilities index (ECI). Displayed for reference only.
maintained by Epoch AI · 5,072 observations on record
Versions
1 version
| Version | Tasks |
|---|---|
| ECI current epoch-capabilities-index | — |
Current leaderboard
ECI · Index points · top 25 of 540 systems
| # | System | Index points |
|---|---|---|
| 1 | Claude Fable 5max effort Anthropic | 162.9 points |
| 2 | Claude Fable 5high effort Anthropic | 162.9 points |
| 3 | Claude Fable 5medium effort Anthropic | 162.9 points |
| 4 | Claude Fable 5xhigh effort Anthropic | 162.9 points |
| 5 | Claude Fable 5 Anthropic | 162.9 points |
| 6 | Claude Fable 5low effort Anthropic | 162.9 points |
| 7 | GPT 5.6 Sol OpenAI | 162.0 points |
| 8 | GPT 5.5 Pro Pre Releasexhigh effort OpenAI | 162.0 points |
| 9 | GPT 5.5 Pro (2026-04-23)high effort OpenAI | 162.0 points |
| 10 | GPT 5.6 Solmax effort OpenAI | 162.0 points |
| 11 | GPT 5.5 Pro (2026-04-23)xhigh effort OpenAI | 162.0 points |
| 12 | GPT 5.6 Sollow effort OpenAI | 162.0 points |
| 13 | GPT 5.5 Pro (2026-04-23) OpenAI | 162.0 points |
| 14 | GPT 5.6 Sol OpenAI | 162.0 points |
| 15 | GPT 5.6 Solmedium effort OpenAI | 162.0 points |
| 16 | GPT 5.6 Solmax effort OpenAI | 162.0 points |
| 17 | GPT 5.6 Solhigh effort OpenAI | 162.0 points |
| 18 | GPT 5.6 Solxhigh effort OpenAI | 162.0 points |
| 19 | GPT 5.5 Pro Pre Releasehigh effort OpenAI | 162.0 points |
| 20 | GPT 5.6 Solno reasoning OpenAI | 162.0 points |
| 21 | GPT 5.5 Pro Unknown OpenAI | 160.7 points |
| 22 | GPT 5.5 (2026-04-22)high effort OpenAI | 159.2 points |
| 23 | GPT 5.5 (2026-04-22)medium effort OpenAI | 159.2 points |
| 24 | GPT 5.5 Pre Releasexhigh effort OpenAI | 159.2 points |
| 25 | GPT 5.5 (2026-04-22)xhigh effort OpenAI | 159.2 points |
One row per evaluated system — reasoning-effort variants rank separately.