Claw-Eval: Toward Trustworthy Evaluation of Autonomous Agents

Benchmark Model Rank Results
agents-on-claw-eval-general-pass-3Qwen3.5-397B-A17B (General)#5Pass^3: 57.8
agents-on-claw-eval-general-pass-3GLM-5V-Turbo (General)#8Pass^3: 52.8
agents-on-claw-eval-general-pass-3DeepSeek-V3.2 (General)#9Pass^3: 42.2
agents-on-claw-eval-general-pass-3NVIDIA-Nemotron-3-Super-120B-A12B-BF16 (General)#10Pass^3: 6.8
agents-on-claw-eval-multi-turn-pass-3GLM-5V-Turbo (Multi Turn)#5Pass^3: 57.9
agents-on-claw-eval-multi-turn-pass-3Qwen3.5-397B-A17B (Multi Turn)#7Pass^3: 52.6