OSWorld 2.0 (Meta, Sep 2026)Agent

The source chart does not identify the task release, scoring mode, or agent harness. This result is kept separate from first-attempt, partial, and strict OSWorld 2.0 variants.

排行数据源

#模型得分

1

Claude Opus 5Anthropic · max 模式

68.3%

2

Muse Spark 1.3Meta · max 模式

66.9%

3

GPT-5.6 SolOpenAI · max 模式

62.7%

4

Muse Spark 1.2Meta · xhigh 模式

47.6%

共 4 个模型

返回基准目录