Unverified50% confidenceBenchmarkExact time
SWE-Bench由普林斯顿大学团队发布，从GitHub上12个真实Python开源项目抽取近2300个真实Issue，2024年初顶级模型通过率仅约4%-5%，至2025年头部模型已突破60%
1
Sources
50%
Confidence
Long-term
Relevance
7/24/2026
First Seen
Sources
Grok 4.5深度解析：专为编程与Agent而生的工程大脑
bilibili傻狗与AI7/9/2026
Related Claims
UnverifiedSWE-Bench was proposed by the Princeton NLP team in 2023, crawling 2,294 real Issue-PR pairs from GitHub and covering 12 mainstream Python projects.83% similarUnverifiedSWE-bench is a large language model programming capability evaluation benchmark introduced by a Princeton University research team in 2023, extracting 2,294 task instances from 12 mainstream Python open-source projects on GitHub.80% similarVerifiedSWE-Bench collected 2,294 real Issue–Pull Request pairs from 12 mainstream Python open-source projects on GitHub.80% similarVerifiedSWE-Bench was introduced by Princeton University in 2023 and evaluates AI programming ability by drawing thousands of resolved issues from real open-source projects on GitHub.79% similarVerifiedSWE-bench was released in 2023 by a research team at Princeton University, collecting 2,294 real GitHub Issues and their corresponding Pull Request fixes from 12 well-known Python open-source projects.79% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/611173API
curl https://kongchang.com/api/v1/knowledge/claims/611173MCP
get_claim(id=611173)