Skip to content

VideoMMMU

A video extension of MMMU-style multimodal reasoning over expert questions grounded in temporal media.

12 models

Higher is better.

  1. 1Qwen3.8 MaxAlibaba88.7
  2. 2Gemini 3 ProGoogle87.6
  3. 3dots3-note PreviewDots Studio86.8
  4. 4Kimi K2.5Moonshot AI86.6
  5. 5GPT-5.2OpenAI85.9
  6. 6Qwen3.7 PlusAlibaba85.4
  7. 7Qwen3.5 397BAlibaba84.7
  8. 8MiniMax M3MiniMax84.6
  9. 9Claude Opus 4.5Anthropic84.4
  10. 10Qwen3.6-27BAlibaba84.4
  11. 11Qwen3.6 PlusAlibaba84.0
  12. 12Qwen3.6-35B-A3BAlibaba83.7
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Video + text reasoning
Tasks
Video-grounded expert reasoning
Authors
Qwen (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.