Skip to content

WideResearch

A broad research-agent benchmark for open-ended information gathering, synthesis, and answer construction across wide search spaces.

16 models

Higher is better.

  1. 1Hy4 previewTencent83.9
  2. 2Ling 3.1 FlashInclusionAI83.3
  3. 3Qwen3.8 MaxAlibaba81.9
  4. 4Atria Dawn PreviewShanghai Artificial Intelligence Laboratory81.9
  5. 5Ornith-1.5-397BOrnith AI80.8
  6. 6Kimi K2.6Moonshot AI80.8
  7. 7dots3-note PreviewDots Studio78.9
  8. 8Claude Opus 4.5Anthropic76.4
  9. 9Qwen3.6 PlusAlibaba74.3
  10. 10Qwen3.5 397BAlibaba74.0
  11. 11Ling 3.0 FlashInclusionAI73.6
  12. 12Kimi K2.5Moonshot AI72.7
  13. 13GLM-5Z.AI69.8
  14. 14Ornith-1.5-35B-A3BOrnith AI67.8
  15. 15Qwen3.6-35B-A3BAlibaba60.1
  16. 16Ornith-1.5-9BOrnith AI59.5
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Multi-source research evaluation
Tasks
Open-ended research tasks
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.