DeepSWE Benchmark

What is DeepSWE?

Evaluates coding agents on 113 original, long-horizon software engineering tasks across 91 open-source repositories in five languages. Tasks are written from…

Released

2026-05-18

Evaluates

Code & Software, Software & AI Compute, cs.SE

Openness

unknown

Importer

Claire Radar

Review status

source-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.07946

DeepSWE paper, code and dataset

• Paper https://arxiv.org/abs/2607.07946
• Code https://github.com/datacurve-ai/deep-swe
• Dataset https://huggingface.co/datasets/datacurve/deep-swe
• Website https://deepswe.datacurve.ai/

Which sources cite DeepSWE?

• DeepSWE

Related Code & Software, Software & AI Compute, cs.SE benchmarks

• Active-SWE
• AFT-Bench
• Agent App Benchmark
• Agent Retrieval Bench
• Agent Systems Benchmark (ASB)
• Agentic Evaluation in 2D Mazes
• Agentic Security Benchmark
• AgentInfraBench