AGI-Eval Community Base - General Reasoning: leaderboard

AGI-Eval Community private-set Base Language Models evaluation (General Reasoning) 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
1ERNIE-4.5-300B-A47B-Base87.97
2Kimi K2 Base86.66#939
3DeepSeek-V3.1-Base85.91
4DeepSeek V3 Base85.42
5GLM-4.5-Base84.94
6DeepSeek-V3.2-Exp-Base84
7Llama 4 Maverick Base82.6
8Qwen2.5-72B-Base82.37
9Llama 3.1 405B Base82.3
10Qwen 2.5 32B (Base)81.33
11Qwen3-14B-Base81.19#1473
12Qwen3-30B-A3B-Base77.55#1475
13Qwen2.5-14B-Base76.25
14Qwen3-8B-Base75.21#1754
15Llama 3.1 70B Base74.63

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-general-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-22.