AGI-Eval Community Base - Cognition: leaderboard

AGI-Eval Community private-set Base Language Models evaluation (Cognition) 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
1Kimi K2 Base86.49#939
2DeepSeek-V3.1-Base86.31
3DeepSeek V3 Base85.43
4DeepSeek-V3.2-Exp-Base85.18
5GLM-4.5-Base83.35
6Llama 3.1 405B Base83.27
7Llama 4 Maverick Base82.81
8Llama 3.1 70B Base81.61
9Qwen 2.5 32B (Base)81.6
10Gemma-3-27B-Base76.66
11Qwen2.5-72B-Base76.35
12Qwen3-30B-A3B-Base76.28#1475
13GLM4-9B-Base76.01
14ERNIE-4.5-300B-A47B-Base74.97
15Llama 4 Scout Base74.9#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-cognition · How It Works · Data refreshed daily, snapshot 2026-11-22.