AI Intelligence
Keep up. Compare. Put it to work.
Loading models data.
Different measures answer different questions. Choose a view to explore.
270 entries · Epoch Capabilities Index
| # | Model | ECI | 95% interval |
|---|---|---|---|
| 1 | Anthropic | 168.83 | 165.3–173.9 |
| 2 | OpenAI | 168.81 | 165.4–174.5 |
| 3 | OpenAI | 168.37 | 165.0–174.4 |
| 4 | Anthropic | 166.49 | 163.5–171.3 |
| 5 | Google DeepMind | 166.45 | 163.2–171.5 |
| 6 | Anthropic | 166.18 | 163.1–171.3 |
| 7 | Anthropic | 163.94 | 161.1–168.2 |
| 8 | OpenAI | 163.73 | 160.9–167.4 |
| 9 | Anthropic | 163.42 | 160.5–168.2 |
| 10 | OpenAI | 162.91 | 160.0–166.6 |
| 11 | OpenAI | 162.79 | 160.2–167.1 |
| 12 | Meta AI | 161.85 | 159.4–165.4 |
| 13 | OpenAI | 160.09 | 157.3–163.3 |
| 14 | OpenAI | 159.68 | 157.3–163.7 |
| 15 | OpenAI | 159.57 | 157.1–162.7 |
| 16 | Anthropic | 159.32 | 156.2–162.5 |
| 17 | Anthropic | 159.01 | 156.9–161.9 |
| 18 | Google DeepMind | 157.97 | 155.9–161.2 |
| 19 | Moonshot | 157.86 | 155.1–160.6 |
| 20 | Google DeepMind | 157.44 | 155.1–161.1 |
| 21 | Alibaba | 157.40 | 155.3–160.3 |
| 22 | OpenAI | 157.31 | 155.4–159.8 |
| 23 | Meta AI | 157.21 | 155.3–159.8 |
| 24 | OpenAI | 157.09 | 153.6–161.4 |
| 25 | Anthropic | 156.88 | 154.1–159.7 |
| 26 | Anthropic | 156.79 | 154.9–159.2 |
| 27 | OpenAI | 156.75 | 154.3–159.4 |
| 28 | xAI | 156.57 | 154.7–159.2 |
| 29 | OpenAI | 156.50 | 154.2–159.3 |
| 30 | OpenAI | 156.33 | 153.9–160.0 |
| 31 | Z.ai (Zhipu AI) | 155.99 | 154.2–158.3 |
| 32 | DeepSeek | 155.93 | 153.7–158.6 |
| 33 | Anthropic | 155.66 | 153.8–158.1 |
| 34 | DeepSeek | 155.64 | 153.7–158.0 |
| 35 | Google DeepMind | 155.34 | 153.3–158.1 |
| 36 | Meta AI | 155.11 | 152.7–158.0 |
| 37 | Alibaba | 155.08 | 153.0–157.2 |
| 38 | Meta AI | 155.05 | 152.8–157.6 |
| 39 | Google DeepMind | 154.88 | 152.9–157.4 |
| 40 | DeepSeek | 154.76 | 152.1–156.9 |
Source: Epoch AI, CC BY 4.0. ECI is a capability index, not a percentage. Rank is ordered by score; confidence intervals show uncertainty. About the data