AI³大模型指数
全球大模型综合榜——智分(分层几何融合)与值分(每美元智分),按版本化分组并标明数据可用情况。
ai3-llm-index/v1.1 · cohort v1.1-20260815 · 数据时间 2026年8月20日
| # | 模型 | 智分 | 值分 | 混合价 | 数据可用率 |
|---|---|---|---|---|---|
| 1 | claude-fable-5 | 99 | — | — | 9% |
| 2 | claude-4-5-opus-high-reasoning | 99 | — | — | 9% |
| 3 | claude-sonnet-4-5 | 99 | — | — | 9% |
| 4 | gemini-3-pro | 99 | — | — | 9% |
| 5 | glm-4-5 | 99 | — | — | 9% |
| 6 | gpt-5-5-high | 99 | — | — | 9% |
| 7 | claude-opus-4-6-high | 98.49 | — | — | 9% |
| 8 | claude-opus-4-6 | 98.23 | — | — | 9% |
| 9 | claude-opus-5-max | 97.97 | — | — | 9% |
| 10 | claude-opus-5-high | 97.71 | — | — | 9% |
| 11 | claude-opus-4-7 | 97.45 | — | — | 9% |
| 12 | qwen3.8-max | 97.2 | — | — | 9% |
| 13 | muse-spark-1.1 | 97.19 | — | — | 9% |
| 14 | claude-opus-4-7-high | 96.94 | — | — | 9% |
| 15 | muse-spark | 96.94 | — | — | 9% |
| 16 | muse-spark-1.2-xhigh | 96.94 | — | — | 9% |
| 17 | gemini-3-1-pro-preview | 96.68 | 50 | $4.5 | 9% |
| 18 | kimi-k3-max | 96.42 | — | — | 9% |
| 19 | gemini-3-pro | 96.16 | — | — | 9% |
| 20 | gpt-5.6-sol-xhigh | 95.91 | — | — | 9% |
| 21 | gpt-5.5-high | 95.39 | — | — | 9% |
| 22 | gemini-3.6-flash-high | 94.88 | — | — | 9% |
| 23 | gpt-5.4-high | 94.87 | — | — | 9% |
| 24 | gpt-5-5 | 94.62 | 17.33 | $11.25 | 9% |
| 25 | gpt-5.2-chat-latest-20260210 | 94.36 | — | — | 9% |
| 26 | gemini-3.5-flash-high | 94.1 | — | — | 9% |
| 27 | qwen3.7-max-preview | 93.84 | — | — | 9% |
| 28 | grok-4.20-beta1 | 93.58 | — | — | 9% |
| 29 | claude-opus-4-8 | 93.33 | 25.5 | $10 | 9% |
| 30 | gpt-5.5-instant | 93.07 | — | — | 9% |
| 31 | gemini-3-flash | 92.81 | — | — | 9% |
| 32 | grok-4.20-beta-0309-reasoning | 92.29 | — | — | 9% |
| 33 | gemini-3.5-flash-medium | 92.04 | — | — | 9% |
| 34 | claude-sonnet-4-6 | 91.78 | 41.83 | $6 | 9% |
| 35 | grok-4.20-multi-agent-beta-0309 | 91.52 | — | — | 9% |
| 36 | glm-5-1 | 91.26 | 9.17 | $10.5 | 9% |
| 37 | glm-5.2-max | 91.01 | — | — | 9% |
| 38 | claude-opus-4-5-20251101 | 90.75 | — | — | 9% |
| 39 | grok-4.5 | 90.49 | — | — | 9% |
| 40 | ernie-5.1 | 90.23 | — | — | 9% |
| 41 | grok-4.1-thinking | 89.97 | — | — | 9% |
| 42 | gemini-2.5-pro | 89.88 | — | — | 36% |
| 43 | qwen3.5-max-preview | 89.72 | — | — | 9% |
| 44 | mimo-v2.5-pro | 89.46 | — | — | 9% |
| 45 | gpt-5.4 | 89.2 | — | — | 9% |
| 46 | o3 | 89.06 | — | — | 36% |
| 47 | claude-opus-4-8-high | 88.97 | — | — | 9% |
| 48 | kimi-k2-6 | 88.94 | — | — | 9% |
| 49 | qwen3.6-max-preview | 88.68 | — | — | 9% |
| 50 | grok-4.1 | 88.43 | — | — | 9% |