Skip to content

OmniDocBench 1.5

A document understanding benchmark used in frontier-model comparison tables to measure extraction and grounded reasoning quality on complex documents.

11 models

Higher is better.

  1. 1Qwen3.8 MaxAlibaba92.1
  2. 2MiniMax M3MiniMax91.6
  3. 3Qwen3.7 PlusAlibaba91.4
  4. 4Qwen3.8-27BAlibaba91.1
  5. 5Qwen3.5 397BAlibaba90.8
  6. 6Qwen3.6-35B-A3BAlibaba89.9
  7. 7Kimi K2.5Moonshot AI88.8
  8. 8Gemini 3 ProGoogle88.5
  9. 9Claude Opus 4.5Anthropic87.7
  10. 10GPT-5.2OpenAI85.7
  11. 11Muse Glimmer 30BMeta75.8
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Document understanding benchmark
Tasks
Document understanding tasks
Authors
OpenAI (2026)

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