AGI-Eval Community - Objective Accuracy (English): leaderboard
Metric: Accuracy (%). Source: agi-eval.cn. 139 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 | 91.58 |
| 2 | Claude Opus 4.6 | 91.56 |
| 3 | Gemini 3.1 Pro (Preview) | 91.54 |
| 4 | Gemini 3.5 Flash | 91.42 |
| 5 | Seed 2.0 Pro | 91.19 |
| 6 | Claude Fable 5 | 91.11 |
| 7 | Kimi K2.6 | 90.99 |
| 8 | Gemini 3 Flash (Preview) | 90.78 |
| 9 | Kimi K3 | 90.74 |
| 10 | Qwen 3.7 Max | 90.68 |
| 11 | GPT-5.5 | 90.65 |
| 12 | Seed 2.1 Pro | 90.63 |
| 13 | Gemini 3 Pro (Preview) | 90.61 |
| 14 | Kimi K2.5 (Thinking) | 90.57 |
| 15 | GLM-5.2 | 90.29 |
Interactive version: theaggregate.ai/benchmark?slug=agi-eval-community-objective-accuracy-english · How It Works · Data refreshed daily, snapshot 2026-09-19.