SWE-bench icon
Developer Tools

SWE-bench

🇨🇳 中文 / 🇬🇧 English — 双语介绍
🇨🇳 中文

SWE-bench 用来评估代码模型修复真实 GitHub issue 的能力,题目取自 Python 开源仓库的真实缺陷及对应测试。官网给出排行榜,分「给完整上下文」与「只给问题描述」两种设置对比,是判断编码 Agent 实际修复水平时被引用最多的公开参照。 🔗 榜单: https://www.swebench.com/ | GitHub Stars: 6.1k+

🇬🇧 English

SWE-bench measures how well coding models fix real GitHub issues, using bugs and their tests taken from Python open-source repositories. The site publishes leaderboards under both full-context and issue-only settings, making it the public yardstick most often cited for coding-agent repair ability. 🔗 Leaderboards: https://www.swebench.com/ | GitHub Stars: 6.1k+

📋 Product Info

Category: Developer Tools
Listed on: 2026-10-21
Official site: https://www.swebench.com
Open source: GitHub repository ↗
Tags: 基准测试,编码模型,评测,开源

🎯 Why This Tool

SWE-bench belongs to the Developer Tools category on hedirbase. Tagged with: 基准测试,编码模型,评测,开源.

SWE-bench measures how well coding models fix real GitHub issues, using bugs and their tests taken from Python open-source repositories. The site publishes leaderboards under both full-context and iss

✓ Manually reviewed 📁 Open source 🌐 Curated for independents
🌎 访问 SWE-bench → 📦 GitHub