List your company

Web Agent Tasks on Mind2Web

60.3Success RateGemini-3-Pro
25.35634.42843.552.572Jun 16, 2026

Evaluation results

MethodLinks
Gemini-3-Pro2026.0660.3
Claude-Opus-4.52026.0657.9
GPT-5.12026.0655.1
Qwen3-Coder-480B-A35B-Instruct2026.0654
GLM-4.72026.0653.7
Kimi-K2-Instruct-09052026.0653.4
Qwen3-235B-A22B-Instruct-25072026.0649
DeepSeek-V3.22026.0647.2
Qwen2.5-Coder-14B-Instruct2026.0642.7
LoopCoder-v2R (number of refinement loops)=42026.0641.4
Qwen2.5-Coder-7B-Instruct2026.0638.4
Seed-Coder-8B-Instruct2026.0638.2
LoopCoder-v2R (number of refinement loops)=12026.0635.3
LoopCoder-v2R (number of refinement loops)=32026.0635.1
LoopCoder-v2R (number of refinement loops)=22026.0634.5
Qwen2.5-Coder-32B-Instruct2026.0632.5
DeepSeek-Coder-V2-Lite-Instruct2026.0626.7