Molmo2 Human Preference - Captioning: leaderboard

Metric: Elo rating (bootstrap median, Bradley-Terry; pairwise human preferences on dense captions of 51 videos). Source: arxiv.org. Saturation forecast: Around February 2027. 21 models tracked.

Top models

#ModelScore
1Gemini 2.5 Pro1148
2GPT-51136
3Gemini 3 Pro1126
4Gemini 2.5 Flash1109
5Qwen 3 VL 8B1105
6GPT-5 Mini1086
7Qwen 3 VL 4B1052
8Molmo2-8B1049
9GLM-4.1V-9B (Thinking)1013
10Claude Sonnet 4.51009

Interactive version: theaggregate.ai/benchmark?slug=molmo2-human-preference-captioning · How It Works · Data refreshed daily, snapshot 2026-09-25.