Claude Sonnet 5.5 · SWE-bench Multilingual(300 problems, 9 programming languages) 90.3% · 附件证据(reviewed-ev55-swe-swe-bench-multilingual)

Claude Sonnet 5.5 的 SWE-bench Multilingual · 300 problems, 9 programming languages 官方出处。Adaptive thinking, max effort · 90.3%。见《Claude Sonnet 5.5 System Card》§8.2, PDF p.110。证据 reviewed-ev55-swe-swe-bench-multilingual。

原文与定位

Claude Sonnet 5.5 · SWE-bench Multilingual(300 problems, 9 programming languages),Adaptive thinking, max effort配置,Pass rate 90.3%(具体成绩公布时间未确认)。见《Claude Sonnet 5.5 System Card》§8.2, PDF p.110。 查看成绩条件

SWE-Bench Pro draws problems from actively-maintained repositories with large, multi-file diffs. Sonnet 5.5 achieved 81.3%. … SWE-Bench Multilingual extends the format to 300 problems across 9 programming languages. Sonnet 5.5 achieved 90.3%. … SWE-Bench Multimodal adds visual context (screenshots, design mockups) to the issue descriptions. Sonnet 5.5 achieved 54.3%.
官方材料
Claude Sonnet 5.5 System Card · PDF 第 110 页
定位
附件 · §8.2, PDF p.110
材料版本
PDF cover dated 2026-09-28; 148 pages; collected 2026-11-10
版本公布
未确认
披露时间
未确认 · 首发/后补未确认
采集时间
历史依据
暂无可靠历史版本依据
引用文本

Claude Sonnet 5.5 · SWE-bench Multilingual(300 problems, 9 programming languages),Adaptive thinking, max effort配置,Pass rate 90.3%(具体成绩公布时间未确认)。官方出处 Claude Sonnet 5.5 System Card(附件,PDF cover dated 2026-09-28; 148 pages; collected 2026-11-10)§8.2, PDF p.110,https://www.anthropic.com/claude-sonnet-5-5-system-card#page=110。Benchlog https://benchlog.lekshome.cc/evidence/claude-sonnet-5-5/reviewed-ev55-swe-swe-bench-multilingual/

关联成绩与条件

Claude Sonnet 5.5 · SWE-bench Multilingual 成绩与配置
指标与成绩配置与评测条件证据关系
Pass rate90.3%原文 90.3% · 单位 %
公布 未确认
Adaptive thinking, max effort
设置
Adaptive thinking enabled; max effort.
未知项
Sampling settings may vary by benchmark; see each result.conditions.
工具
工具条件未记录
推理预算
max
框架
未记录
协议
SWE-Bench Multilingual, 300 problems across 9 programming languages; five-trial average.
条件
No further harness details stated in this section.

直接支持该成绩

在排名中查看此成绩