Codesota · Models · UFO (GPT-4V)Microsoft1 results · 1 benchmarks
Model card

UFO (GPT-4V).

Microsoftopen-sourceUnknown paramsUI-Focused agent with dual-agent architecture on GPT-4V

UFO: UI-Focused Windows agent. Dual-agent: AppAgent + ActAgent. arxiv:2402.07939.

§ 02 · Benchmarks

Every benchmark UFO (GPT-4V) has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01OSWorldAgentic AI · Web & Desktop Agentssuccess-rate9.4%#25/282024-04-11source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Strengths by area

Where UFO (GPT-4V) actually performs.

Agentic AI
1
benchmark
avg rank #25.0
§ 04 · Papers

1 paper with results for UFO (GPT-4V).

  1. 2024-04-11· Agentic AI· 1 result

    OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments

§ 05 · Related models

Other Microsoft models scored on Codesota.

Phi-4
14B params · 17 results
Phi-4 Multimodal Instruct
6B params · 9 results
WizardLM-2-8x22b
7 results
DeBERTa-v3-large
304M params · 6 results
E5-Mistral-7B-instruct
7B params · 4 results
ResNet-152
60M params · 3 results
ResNet-50
25M params · 3 results
UniXcoder
Unknown params · 3 results
§ 06 · Sources & freshness

Where these numbers come from.

arxiv
1
result
1 of 1 rows marked verified.