Unverified70% confidenceFactTime unknown
人类在OS World测试中的平均得分约为72-74%
1
Sources
70%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedResearch shows that the agreement rate between LLM judges and human scores is approximately 80%, approaching but not yet surpassing the level of human inter-annotator agreement (approximately 81%).62% similarUnverifiedThe current strongest models score approximately 78% on GUI operations, making Browser Use and Computer Use viable only as fallback solutions.59% similarUnverifiedSol achieved 92.2% on the web browsing benchmark BrowseComp and 62.6% on OSWorld 2.0, both current best results; on OSWorld it surpassed Opus 4.8 while reducing output tokens by 85%.59% similarUnverified在Terminal Bench 2.1基准测试上，Ultra模式接近92%，超越竞品的88%表现58% similarUnverifiedAnthropic的Computer Use在OSWorld基准测试上的操作准确率约为14.9%（2024年10月数据）57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/13718API
curl https://kongchang.com/api/v1/knowledge/claims/13718MCP
get_claim(id=13718)