AGI-Eval Community - Cognition (English): leaderboard
AGI-Eval community leaderboard (agi-eval.cn), the private-set LLM board built from the platform's non-public items; latest monthly snapshot. Unrelated to the AGIEval exam dataset. This board is the English-language item split (_英) of the cognition column; the parent column combines both splits. Upstream accuracy 0-1, shown as 0-100; higher is better.
Metric: Accuracy (%). Source: agi-eval.cn. Saturation forecast: Estimated already saturated. 139 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4 (Thinking) | 89.88 | #170 (Claude Opus 4) |
| 2 | Claude Opus 4.6 | 88.41 | #67 |
| 3 | Grok 4 | 87.63 | #189 |
| 4 | Claude Opus 4.5 (Non-reasoning) | 87.19 | #87 (Claude Opus 4.5) |
| 5 | Gemini 3 Flash (Preview) | 87.06 | #90 |
| 6 | Gemini 3.1 Pro (Preview) | 86.97 | #64 |
| 7 | Gemini 3.5 Flash | 86.85 | #66 |
| 8 | Claude Fable 5 | 86.78 | #20 |
| 9 | Seed 1.8 | 86.63 | #143 |
| 10 | Claude Sonnet 4 (Thinking) | 86.34 | #207 (Claude Sonnet 4) |
| 11 | Seed 2.0 Pro | 86.05 | #122 |
| 12 | Kimi K2.6 | 85.74 | #113 |
| 13 | Claude Opus 4.8 | 85.69 | #39 |
| 14 | Doubao-Seed-1.6 (Thinking) | 85.56 | |
| 15 | O3 Pro | 85.56 | #116 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agi-eval-community-cognition-english · How It Works · Data refreshed daily, snapshot 2026-11-05.