AGI-Eval Community - Interaction (Chinese): 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 Chinese-language item split (_中) of the interaction 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 4.8 | 89.93 | #39 |
| 2 | Claude Fable 5 | 89.52 | #20 |
| 3 | GPT-5.5 | 89.4 | #33 |
| 4 | GPT-5.4 | 88.45 | #83 |
| 5 | Claude Opus 4.6 | 88.39 | #67 |
| 6 | Seed 2.0 Pro | 88.35 | #122 |
| 7 | Gemini 3.1 Pro (Preview) | 87.99 | #64 |
| 8 | Kimi K3 | 87.74 | #35 |
| 9 | Gemini 3.5 Flash | 87.45 | #66 |
| 10 | Seed 2.1 Pro | 87.31 | #73 |
| 11 | Gemini 3 Flash (Preview) | 86.86 | #90 |
| 12 | Qwen 3.7 Max | 86.71 | #85 |
| 13 | O3 Pro | 85.87 | #116 |
| 14 | Claude Sonnet 5 | 85.67 | #65 |
| 15 | Kimi K2.6 | 85.28 | #113 |
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-interaction-chinese · How It Works · Data refreshed daily, snapshot 2026-11-05.