OSWorld
[NeurIPS 2024] OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
OSWorld scores 6.1/10 (Solid) in Benchmarks & Resources.
Kind Resource Language Python License Apache-2.0 Stars 3k★ Repo xlang-ai/osworld
Quality
Solid 6.1
Adoption 37
Activity 62
Maturity 95
Community 43
Capability 50
Show the math
Overall 6.1/10
value = 6.1/10
score = value
Adoption 4.3/10
value = 36.9/100
score = 1 + 9 * value / 100
Activity 6.6/10
value = 62.2/100
score = 1 + 9 * value / 100
Maturity 9.5/10
value = 94.7/100
score = 1 + 9 * value / 100
Community 4.9/10
value = 42.9/100
score = 1 + 9 * value / 100
Capability 5.5/10
value = 50/100
score = 1 + 9 * value / 100
Key metrics
3k★ GitHub stars
22/90d Recent commits
18 devs/90d Recent contributors
2.7y Project age
active Last commit
Gotchas
No gotchas documented yet. Contribute one if you know a constraint we missed.
benchmark computer-use multimodal agents