List your company

Web Navigation Success on VisualWebArena

19.8Success RateGPT-4o
2.2246.78711.3515.913May 29, 2025

Evaluation results

MethodLinks
GPT-4oInput Modality=with text2025.0519.8
ViGoRL-7bBackbone=7b2025.0511.2
ICAL-7BBackbone=7B, Input Modality=with text2025.058.2
ViGoRL-3bBackbone=3b2025.056.4
Qwen2.5-VL-7BBackbone=7B2025.055.5
Qwen2.5-VL-3B + Vanilla GRPOBackbone=3B, Training Protocol=Vanilla GRPO2025.054.8
Qwen2.5-VL-3B + SFT directBackbone=3B, Training Protocol=SFT direct2025.054.5
Qwen2.5-VL-3BBackbone=3B2025.054.2
Qwen2-VL-7BBackbone=7B, Input Modality=with text2025.052.9