List your company

Web Navigation on Multimodal-Mind2Web Cross-Domain

57.4Element AccuracySkill-CMIB
12.47224.13635.847.464May 8, 2026

Evaluation results

MethodLinks
Skill-CMIBLearning Paradigm=Skill Based2026.0557.469.542.9
AgentTrek-7BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.055687.552.6
Explorer-7BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.0555.790.453
Explorer-7BLearning Paradigm=Supervised Fine-Tuning2026.0550.988.948.1
Explorer-4BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.0549.888.847.2
Explorer-4BLearning Paradigm=Supervised Fine-Tuning2026.0546.987.744.6
Qwen2.5-VL-7B-Inst.Learning Paradigm=Skill Based2026.0546.866.332.4
SeeActLearning Paradigm=In-Context Learning2026.0542.469.336.8
Scribe-Agent-L.-32BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.0539.454.737.3
GPT-4Learning Paradigm=In-Context Learning2026.0535.461.929.7
GPT-3.5Learning Paradigm=In-Context Learning2026.0525.257.924.1
ClassificationLearning Paradigm=Supervised Fine-Tuning2026.0524.5——
SeeClick-9.6BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.0522.184.120.2
MiniCPM-GUI-3.1BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.0517.974.514.6
GenerationLearning Paradigm=Supervised Fine-Tuning2026.0514.244.711.9
EDGE-9.6BLearning Paradigm=Data Synthesis+Supervised Fine-Tuning2026.05——22.4