Skip to content

MLVU mean average

A multi-task video understanding benchmark averaged across MLVU categories.

9 models

Higher is better.

  1. 1Qwen3.8 MaxAlibaba90.8
  2. 2Qwen3.7 PlusAlibaba87.4
  3. 3Qwen3.5 397BAlibaba86.7
  4. 4Qwen3.6-27BAlibaba86.6
  5. 5Qwen3.6-35B-A3BAlibaba86.2
  6. 6GPT-5.2OpenAI85.6
  7. 7Kimi K2.5Moonshot AI85.0
  8. 8Gemini 3 ProGoogle83.0
  9. 9Claude Opus 4.5Anthropic81.7
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Video QA and understanding
Tasks
General video understanding
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.