A compact, text-focused model built around coding tasks, operating at the 4B parameter scale where efficiency meets utility. It handles code generation and related text tasks within a small footprint, making it practical for local deployment. As a community-published variant, it sits outside official release channels, so behavior and quality may vary from the base Qwen3.5 family.