TwinLab-Bench - Mean Score: leaderboard

TwinLab-Bench (arXiv 2611.26248) scientific experimentation benchmark (Mean Score) across 52 tasks in agentic twins of real laboratories.

Metric: Mean Rubric Score (0-100 across 52 tasks). Source: huggingface.co. Saturation forecast: Around May 2028. 13 models tracked. Updated 22 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Fable 5.2High45#2
2Claude Opus 5.5High40.2#9
3GPT-6.1 SolHigh31.7#15
4GPT-6 AstraHigh30.7#11
5Claude Sonnet 5.5High28.7#24
6DeepSeek V4.1 Flash (1017)High20#48
7GLM-5.3High19.4#79
8Kimi K3.1 ProHigh18.1#35
9MiMo-V2.6-ProHigh17.9#94
10Qwen 3.8 MaxHigh13.1#67
11hy4High12.9#44
12MiMo-V2.6-FlashHigh10.9#183
13GPT-6 LunaHigh10.6#108

No result here: #3 Gemini 4 Neon, #4 GPT-6.2 Astra, #7 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=twinlab-bench-mean-score · How It Works · Data refreshed daily, snapshot 2026-11-22.