OpenHands Index - Commit0: leaderboard

OpenHands Index Commit0 slice measures agent test-pass performance on repository commit tasks.

Metric: Score (%). Source: huggingface.co. 5 models tracked.

Top models

#ModelScoreOverall rank
1OpenHands CodeAct v2.1 (claude-3-5-sonnet-20241022)71.2
2OpenHands CodeAct v2.0 (gpt-4o-2024-11-20)68.9
3AutoCodeRover (gpt-4-turbo-2024-04-09)61.5
4Agentless (gpt-4o-mini-2024-07-18)55.3
5SWE-Agent (claude-3-opus-20240229)52.1

Interactive version: theaggregate.ai/benchmark?slug=openhands-index-commit0 · How It Works · Data refreshed daily, snapshot 2026-10-20.