Can agents really use computers? a16z answers with data
Original title:Can Agents Use a Computer Yet? We've Got the Data
AI summary
a16z data shows that the best score for computer-use agents on the OSWorld-Verified benchmark has risen from 42% a year ago to 85%, surpassing the roughly 72% level of human testers, with Claude Fable 5 leading at 85%.
Key Takeaways
- a16z data shows that the best score for computer-use agents on the OSWorld-Verified benchmark has…
- surpassing the roughly 72% level of human testers
- with Claude Fable 5 leading at 85%.
Quick Facts
- Published
- Updated
- Source
- a16z:News(RSS)
- Category
- Opinion
- AI score
- 72
用 OSWorld 基准和一线案例回答了计算机使用代理能否可靠运行,更重要的判断是基础模型正在成为可替换层,真正的壁垒在于上下文、验证和故障处理。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Can agents really use computers? a16z answers with data. August 10, 2026. https://new.nat6.net/en/news/cmsnbzkyr01qjrohfmjqlmjk5/