AGI-Eval Community Base - Mathematical Reasoning (English): leaderboard

AGI-Eval Community private-set Base Language Models evaluation (Mathematical Reasoning (English)) 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-Base74.67
2Kimi K2 Base69.15#939
3DeepSeek-V3.2-Exp-Base65.66
4DeepSeek-V3.1-Base65.25
5DeepSeek V3 Base64.42
6Qwen3-30B-A3B-Base63.7#1475
7Qwen2.5-72B-Base63.59
8GLM-4.5-Base63.33
9Llama 4 Maverick Base62.63
10Qwen 2.5 32B (Base)62.33
11Qwen3-14B-Base62.04#1473
12Qwen3-8B-Base56.03#1754
13Qwen2.5-14B-Base54.99
14Qwen3-4B-Base53.01#2439
15Llama 3.1 405B Base52.27

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