AGI-Eval Community Base - Interaction (English): leaderboard

AGI-Eval Community private-set Base Language Models evaluation (Interaction (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
1Qwen2.5-72B-Base73.43
2Llama 3.1 405B Base73.18
3DeepSeek-V3.1-Base73.13
4Qwen3-14B-Base72.38#1473
5Qwen 2.5 32B (Base)71.68
6DeepSeek V3 Base71.3
7DeepSeek-V3.2-Exp-Base70.07
8GLM4-9B-Base68.67
9ERNIE-4.5-300B-A47B-Base68.34
10Kimi K2 Base67.71#939
11Gemma-3-12B-Base67.18
12Qwen2.5-14B-Base66.69
13Qwen3-8B-Base66.23#1754
14Qwen 2.5 7B (Base)66.2
15Gemma-3-4B-Base65.94

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