Cognition Releases SWE-1.7 Programming Model Reaching 1000 Tokens Per Second

According to Beating, Cognition released its new AI programming model SWE-1.7, now integrated into Devin. Built on Kimi K2.7 Code by Moon Dimension, SWE-1.7 performs near GPT-5.5 and Claude Opus 4.8 levels while significantly reducing generation costs. The model achieves 1000 tokens per second throughput by integrating Cerebras inference chips. The model also extends task duration to up to 6 hours by automatically summarizing context at critical points and continuing from summaries.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments