DeepSWE benchmark shows big gap between closed and open models

View on Reddit

A Reddit post highlights a new benchmark, DeepSWE, revealing a large performance gap between proprietary and open-source AI models. The post includes a comparison chart. Commenters express hope that open-source will catch up.