Cognition Releases SWE-2 Flash, Beating Larger US Open Models

Cognition launched SWE-2 Flash, a 276B-A12B MoE coding model trained in just 6 weeks, which it says outperforms all US open-weight models, including ones 3x its size.

2026-11-05 ~ 2026-11-05 · 3 related posts

• Cognition launches SWE-2 Flash, a 276B-A12B coding model that beats all US open-weight rivals — carlesgelada · 2026-11-05
• Cognition launches SWE-2 Flash, beating US open-weight models 3x its size after 6 weeks — simonguozirui · 2026-11-05

1 near-duplicate retellings: mishig25