Skip to content

QwenClawBench

Qwen's internal OpenClaw-style benchmark for measuring broad real-world agent performance across practical productivity and research tasks.

10 models

Higher is better.

  1. 1Qwen3.7 MaxAlibaba64.3
  2. 2Qwen3.7 PlusAlibaba61.8
  3. 3Qwen 3.6 Max (preview)Alibaba59.0
  4. 4Qwen3.6 PlusAlibaba57.2
  5. 5Kimi K2.5Moonshot AI54.3
  6. 6GLM-5Z.AI54.1
  7. 7Qwen3.6-27BAlibaba53.4
  8. 8Qwen3.6-35B-A3BAlibaba52.6
  9. 9Claude Opus 4.5Anthropic52.3
  10. 10Qwen3.5 397BAlibaba51.8
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
End-to-end agent evaluation
Tasks
Real-world agent workflows
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.