Cognition launches SWE-1.7: near GPT-5.5 coding intelligence trained from Kimi K2.7 at a fraction of the cost

SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence

Cognition released SWE-1.7, a coding model trained via RL post-training on a Kimi K2.7 base. It scores 42.3% on FrontierCode 1.1, close to GPT-5.5’s 43.0% and a huge jump from SWE-1.6’s 9.4%. It also hits 81.5% on Terminal-Bench 2.1 and 77.8% on SWE-Bench Multilingual, both competitive with GPT-5.5. The gains come from four RL pipeline upgrades: top-p sampling with distribution replay to prevent entropy collapse, multi-continent multi-cluster training with fault tolerance, automated execution-based data filtering, and self-compaction that lets the model summarize long-horizon task state to exceed the context window. SWE-1.7 is live in Devin via Cerebras at 1000 TPS. The post does not disclose specific pricing, only that it advances the cost-performance curve.

Why it matters: Cognition drops SWE-1.7: RL post-training on a Kimi K2.7 base lifts FrontierCode 1.1 pass rate from 9.4% to 42.3%, within a point of GPT-5.5. The numbers are solid and the narrative is sharp, but the post only gives a summary—training details and cost comparisons aren't spelle...