Skip to content
AIMarketCap

SWE-1.7

Model

Agentic coding model optimized for long-horizon software tasks and frontier cost-performance, trained from a Kimi K2.7 base.

Released
Jul 8, 2026
Context window
—
Modality
—
Open source
—
API available
—
Releases
1

Releases

1
Jul 8, 2026Cognition launches SWE-1.7SWE-1.7 · SWE-1.7

Cognition released SWE-1.7, its then-most-capable coding model, optimized for long-horizon agentic tasks and served at up to 1,000 tokens per second.