SWE-1.7

Model for long-horizon asynchronous coding: Cognition's reinforcement fine-tune on top of Moonshot AI's open-weight Kimi K2.7 Code. It scores 42.3% on FrontierCode 1.1 against SWE-1.6's 9.4% and runs in Devin (web, desktop, CLI) on Cerebras at about 1000 tokens per second.

🔴 производитель по правилу «кто обучил» базовую модель — Moonshot AI; Cognition сделала дообучение. Анонс 08.07.2026

Understands

text

This model cannot see images or video — you can only give it text. That matters if you planned to show it screenshots or documents with pictures.

Produces

text, code

Contents verified with the vendor: 2026-08-18

Included in subscriptions

No catalogue entry marks this model as part of a plan yet. It will appear here once the plan contents are confirmed with the vendor.