🤖 Agentic · source: llm-stats.com

OSWorld

Real computer-use tasks across operating systems: browsers, office apps and file management.

2
Models Tested
Claude Opus 4.6
Top Model
72.7%
Top Score
Anthropic
Provider
#ModelOSWorldProviderPodium Score
1
72.7%
Anthropic61.2
2
72.5%
Anthropic38.1
Benchmark data aggregated from llm-stats.com. See Methodology for details.