AGI-Eval Community Base - Objective Accuracy (Chinese): leaderboard
AGI-Eval Community private-set Base Language Models evaluation (Objective Accuracy (Chinese)) measuring accuracy (%) across bilingual capability dimensions.
Metric: Accuracy (%). Source: agi-eval.org. Saturation forecast: Estimated already saturated. 25 models tracked. Updated 22 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2 Base | 76.8 | #939 |
| 2 | ERNIE-4.5-300B-A47B-Base | 72.51 | |
| 3 | DeepSeek-V3.1-Base | 71.62 | |
| 4 | DeepSeek-V3.2-Exp-Base | 70.87 | |
| 5 | GLM-4.5-Base | 69.83 | |
| 6 | DeepSeek V3 Base | 69.75 | |
| 7 | Llama 4 Maverick Base | 68.06 | |
| 8 | Qwen 2.5 32B (Base) | 66.33 | |
| 9 | Qwen2.5-72B-Base | 65.95 | |
| 10 | Qwen3-30B-A3B-Base | 64.53 | #1475 |
| 11 | Qwen3-14B-Base | 61.94 | #1473 |
| 12 | Qwen3-8B-Base | 60.11 | #1754 |
| 13 | Qwen2.5-14B-Base | 59.82 | |
| 14 | Llama 3.1 405B Base | 59.78 | |
| 15 | Qwen3-4B-Base | 57 | #2439 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=agi-eval-community-base-objective-accuracy-chinese · How It Works · Data refreshed daily, snapshot 2026-11-22.