Cognition releases SWE-1.7 coding agent, reaching near GPT-5.5 and Opus intelligence levels, Hacker News reports

Cognition's SWE-1.7 achieves performance comparable to GPT-5.5 and Opus in coding tasks. The release marks a significant advance in AI coding agents, reaching near frontier model intelligence.

• SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence