MMVU

reasoning

MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.

Leaderboard

Showing 4 of 4 results

Kimi K2.5

80.4%

i
Qwen3.5-122B-A10B

74.7%

i
Qwen3.5-27B

73.3%

i
Qwen3.5-35B-A3B

72.3%

i