SWE-2 Flash Released
The story so far
qwen released the low-cost intelligent coding model SWE-2 Flash in the US, introducing a mid-training stage before reinforcement learning for the first time, significantly outperforming baseline models Inkling-Small and FrontierCode 2.0. Scores on Terminal-Bench 4.0 and FrontierCode 2.0 increased from 0.9% to 20.0% and from 3.2% to 24.0%, respectively, surpassing all US open-source models. The model is available in Devin Desktop and CLI, priced at one-fifth of SWE-2 and three times faster.
AI-written from the reports · Updated 14 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- Cognition BlogIntroducing SWE-2 Flash
SWE-2 Flash introduces a mid-training stage before reinforcement learning for the first time, significantly outperforming the baseline model Inkling-Small. Terminal-Bench 4.0 scores increased from 0.9% to 20.0%, and FrontierCode 2.0 improved from 3.2% to 24.0%, surpassing all US open-source models. This model is available in Devin Desktop and CLI, priced at one-fifth of SWE-2 and three times faster.