List your company

Next-step Generation on Human Evaluation Subset

65Win RateCoP (Qwen2.5-VL-7B)
47.3251.9156.561.09May 14, 2026

Evaluation results

MethodLinks
CoP (Qwen2.5-VL-7B)Backbone=Qwen2.5-VL-7B2026.0565233
CoP (GPT-4o)Backbone=GPT-4o2026.05482031