List your company

Computer Use on OSWorld

75OS Success RateEvoCUA-32B
-0.483219.113438.7158.3066Dec 20, 2024Mar 17, 2025Jun 12, 2025Sep 7, 2025Dec 3, 2025Feb 28, 2026May 27, 2026

Evaluation results

MethodLinks
EvoCUA-32B2026.057551.0652.98————65.22——61.07————
R-WoMBackbone=Claude-4.5-Sonnet2025.1067.84——————————————
WoMBackbone=Claude-4.5-Sonnet2025.1062.09——————————————
RAGBackbone=Claude-4.5-Sonnet2025.1060.43——————————————
VanillaBackbone=Claude-4.5-Sonnet2025.1059.12——————————————
R-WoMBackbone=Claude-4-Sonnet2025.1056.73——————————————
RAGBackbone=Claude-4-Sonnet2025.1050.82——————————————
WoMBackbone=Claude-4-Sonnet2025.1049.65——————————————
R-WoMBackbone Model=Claude-4.5-Sonnet, Setting=tutorial-scarce2025.1049.29——————————————
VanillaBackbone=Claude-4-Sonnet2025.1048.24——————————————
WebDreamerBackbone Model=Claude-4.5-Sonnet, Setting=tutorial-scarce2025.1046.35——————————————
RAGBackbone Model=Claude-4.5-Sonnet, Setting=tutorial-scarce2025.1046.11——————————————
VanillaBackbone Model=Claude-4.5-Sonnet, Setting=tutorial-scarce2025.1045.83——————————————
No DefenseAgent=Claude Sonnet 4.52026.0242.9——————————————
GPT-4o + AllyInput Modalities=Accessibility tree (A11y) and screenshot, Planner=GPT-4o, Grounding=Ally2024.1241.674.266.818.79.56.6715.2230.4307.4611.21————
DEACTIONAgent=Claude Sonnet 4.52026.0240.7——————————————
DEACTIONAgent=OpenCUA-72B2026.0239.6——————————————
R-WoMBackbone Model=Claude-4-Sonnet, Setting=tutorial-scarce2025.1039.28——————————————
No DefenseAgent=OpenCUA-72B2026.0239——————————————
DSPAgent=OpenCUA-72B2026.0238.3——————————————
RAGBackbone Model=Claude-4-Sonnet, Setting=tutorial-scarce2025.1036.93——————————————
Task ShieldAgent=OpenCUA-72B2026.0236.6——————————————
InferActAgent=OpenCUA-72B2026.0236.5——————————————
VanillaBackbone Model=Claude-4-Sonnet, Setting=tutorial-scarce2025.1035.82——————————————
R-WoMBackbone Model=Claude-3.7-Sonnet, Setting=tutorial-scarce2025.1035.71——————————————
PromptArmorAgent=OpenCUA-72B2026.0235.2——————————————
WebDreamerBackbone Model=Claude-4-Sonnet, Setting=tutorial-scarce2025.1034.43——————————————
RAGBackbone Model=Claude-3.7-Sonnet, Setting=tutorial-scarce2025.1033.36——————————————
UI-TARS-1.5-7B + LEARNWEAK-GENframework=LEARNWEAK-GEN2026.0533.338.5122.7————40.57——26.28————
VanillaBackbone Model=Claude-3.7-Sonnet, Setting=tutorial-scarce2025.1032.25——————————————
WebDreamerBackbone Model=Claude-3.7-Sonnet, Setting=tutorial-scarce2025.1030.86——————————————
DEACTIONAgent=OpenAI CUA2026.0230.7——————————————
No DefenseAgent=OpenAI CUA2026.0226——————————————
GPT-4o + Aria-UI_THInput Modalities=Pure-vision (screenshots only), Planner=GPT-4o, Grounding=Aria-UI_TH2024.12254.2615.328.730.0626.6723.821.7419.238.5515.15————
GPT-4o + SoMInput Modalities=Accessibility tree (A11y) and screenshot, Planner=GPT-4o, Grounding=SoM2024.1220.8306.774.356.5304.354.3503.64.59————
UI-TARS-1.5-7B2026.0516.677.0921.98————30.43——19.04————
Claude2025.0214.9——————————————
Aguvis-72B2025.0210.26——————————————
Qwen2.5-VL-72B2025.028.83——————————————
GPT-4oInput Modalities=Pure-vision (screenshots only)2024.128.3306.774.3516.104.354.353.855.585.03————
GPT-4o2025.025.03——————————————
Gemini 2.02025.024.7——————————————
CogAgent + SoMInput Modalities=Accessibility tree (A11y) and screenshot, Planner=CogAgent, Grounding=SoM2024.124.172.1704.346.5302.170000.99————
CogAgentInput Modalities=Pure-vision (screenshots only)2024.124.172.1704.356.5302.17000.11.11————
Qwen2-VL-72B2025.022.42——————————————
Claude-Sonnet-4.52026.03———————————62.9———
Gemini-2.5-Pro2026.03———————————13.3———
GPT-O3-CUA2026.03———————————38.1———
MACUSetup=Multi-Agent, Manager=Qwen2.5-72B, Subagent backbone=Qwen2.5-72B2026.06——————————44.9—1.114.2-12.4
MACUSetup=Multi-Agent, Manager=Opus 3.5, Subagent backbone=Qwen2.5-72B2026.06——————————48.5—4.721.4-5.2
Seed1.5-VL2026.03———————————36.7———
Seed1.82026.03———————————61.9———
Single-AgentSetup=Single-Agent, Manager=—, Subagent backbone=Qwen2.5-72B2026.06——————————43.8——26.6—