AGI-Eval Community Base - Learning (Chinese): leaderboard

AGI-Eval Community private-set Base Language Models evaluation (Learning (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

#ModelScoreOverall rank
1DeepSeek-V3.1-Base72.78
2Kimi K2 Base70.29#939
3DeepSeek V3 Base70.16
4GLM-4.5-Base68.59
5DeepSeek-V3.2-Exp-Base68.27
6ERNIE-4.5-300B-A47B-Base67.17
7Llama 4 Maverick Base66.86
8Llama 3.1 405B Base66.56
9Qwen 2.5 32B (Base)64.96
10Qwen2.5-72B-Base62.95
11Gemma-3-27B-Base61.35
12Llama 3.1 70B Base60.95
13Qwen3-14B-Base60.33#1473
14Qwen3-30B-A3B-Base59.63#1475
15Llama 4 Scout Base57.72#1967

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-learning-chinese · How It Works · Data refreshed daily, snapshot 2026-11-22.