List your company

GUI agent success on WebVoyager (test)

90.6Success RateGPT-5 (SoM)
47.33658.56869.881.032May 14, 2026

Evaluation results

MethodLinks
GPT-5 (SoM)# Steps=1002026.0590.6
Gemini computer-use-preview# Steps=1002026.0588.6
Gemini-3-flash (Axtree)# Steps=1002026.0585.6
o3 (SoM)# Steps=1002026.0579.3
MolmoWeb-8B# Steps=100, # Tasks=>278.5K2026.0578.2
MolmoWeb-4B# Steps=100, # Tasks=>278.5k2026.0575.2
Gemini-3-flash (Axtree)# Steps=302026.0574.4
Orchard-GUI-4B (SFT + RL)# Steps=30, # Tasks=2.6k2026.0574.1
Fara-7B# Steps=100, # Tasks=>123.2k2026.0573.5
OpenAI computer-use-preview# Steps=1002026.0570.9
GPT-5 (Axtree)# Steps=302026.0570.6
GLM-4.1V-9B-Thinking# Steps=1002026.0566.8
UI-TARS-1.5-7B# Steps=1002026.0566.4
GPT-4o (SoM)# Steps=1002026.0565.1
Qwen3-VL-235B-A22B-Thinking# Steps=302026.0563.1
Orchard-GUI-4B-SFT# Steps=30, # Tasks=0.4k2026.0560.2
Holo1-7B# Steps=30, # Tasks=>15.6k2026.0555.4
Qwen3-VL-4B-Thinking# Steps=302026.0549