Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Fine-Grained Action Execution on OSExpert-Eval
Loading...
28
GIMP Success Rate
OSExpert (Explore w/ Qwen-3-VL-8B)
-1.12
6.44
14
21.56
Mar 9, 2026
GIMP Success Rate
LibreOffice Success Rate
Updated 4mo ago
Evaluation Results
Method
Method
Links
GIMP Success Rate
LibreOffice Success Rate
OSExpert (Explore w/ Qwen-3-VL-8B)
Agent Type=Agentic Fra...
2026.03
28
26
OSExpert (Explore w/ GPT-5 & UI-TARS)
Agent Type=Agentic Fra...
2026.03
28
26
OSExpert (Explore w/ Qwen-3-VL-8B, wo/ Fine Skills)
Agent Type=Agentic Fra...
2026.03
14
10
OpenCUA-7B
Agent Type=Specialized...
2026.03
7
5
Qwen-3-VL-8B
Agent Type=General Model
2026.03
7
10
Computer-Use-Preview
Agent Type=Specialized...
2026.03
0
0
CoAct-1
Agent Type=Agentic Fra...
2026.03
0
5
Agent-S3 w/ GPT-5
Agent Type=Agentic Fra...
2026.03
0
10
Feedback
Search any
task
Search any
task