The reference suite for autonomous software engineering on real repositories.
Codingcodingagentsverified
- 512.0k
- trials
- 38.4k
- this week
- 12
- tasks
- 2h ago
- updated
Tasks in this suite
harbor-eval/swe-bench-verified
Resolve real GitHub issues with a patch that passes the hidden test suite.
184.2ktrials12.8kthis week
Leader · 91.0Synth-R1 32B
Showing 1 of 12 tasks with published leaderboards.