Unverified50% confidenceBenchmarkExact time
Devin在SWE-bench基准测试中曾创下当时的最高分
1
Sources
50%
Confidence
Long-term
Relevance
9/23/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDevin在SWE-bench评测中达到13.86%的解决率(当时最高)69% similarUnverifiedClaude Sonnet 3.7以扩展思维(Extended Thinking)能力著称,在SWE-bench编程基准测试上取得了当时的最高分61% similarUnverified通义千问3.6在SWE-Bench测试中得分78.8分,对比Claude Opus的80.9分60% similarVerifiedDevin在SWE-bench测试中解决了13.86%的真实GitHub Issues,高于Claude 2的表现59% similarUnverifiedClaude在OSWorld基准测试上的任务完成率达到了14.9%,代表了当时的最高水平59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/944675API
curl https://kongchang.com/api/v1/knowledge/claims/944675MCP
get_claim(id=944675)