Skip to content

Video-MME

A comprehensive benchmark for multimodal large language models on video understanding, covering temporal reasoning, perception, and question answering over videos.

2 models

Higher is better.

  1. 1Seed 2.1 ProByteDance89.2
  2. 2Kimi K2.5Moonshot AI87.4
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Video QA and analysis
Tasks
Video understanding
Authors
Video-MME benchmark team (2024)

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