Cognition launches SWE-2 coding model, pushing the cost-performance frontier

Cognition 发布 SWE-2 编码模型，以更低成本逼近前沿

Cognition's SWE-2 hits 50.0% on FrontierCode 1.1 Main, 8 points above SWE-1.7, at 64% lower cost than Fable 5.1. It's post-trained from the 2.8T-parameter Kimi K3 using a single-run RL method that trains all reasoning-effort levels together. The medium effort level solves tasks in 53 steps vs. 127 for SWE-1.7. The post doesn't disclose exact API pricing.

Why it matters: SWE-2 hits 50.0% on FrontierCode 1.1 Main, +8 pts over its predecessor, and costs 64% less than Fable 5.1 — Cognition's closest model to the frontier yet. Not an 85 because it still trails GPT-6 Astra by a few points and uses Kimi K3 as the base rather than an in-house model, ...