Unverified50% confidenceBenchmarkExact time
In the OSWorld V2.1 benchmark, Muse Wonder Full achieved 78.4%, falling short of Claude Opus 5.5's 81.8%, and Anthropic still has a marginal advantage in desktop environment control.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/13/2026
First Seen
Valid until: 10/27/2026
Sources
Related Entities
Related Claims
UnverifiedAnthropic ran Claude Fable 5.1 on OSWorld 2.0 five times and averaged the results, and Yutori ran its own Navigator n2 three times.58% similarUnverifiedMuse Wonder Full scores 59.1 on Meta's Personal Superintelligence Index, leading all models.57% similarVerifiedOn Automation Bench, Gemini 4 Argon scored 51.3%, surpassing Fable 5.1's 31.4% and Opus 5.5's 42.5%.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1020383API
curl https://kongchang.com/api/v1/knowledge/claims/1020383MCP
get_claim(id=1020383)