A compact but surprisingly capable text model that punches above its weight class for its size. Qwen3 4B handles reasoning and instruction-following with reasonable fluency, though it naturally shows limitations on complex multi-step tasks compared to larger siblings. Its open-weight nature and GGUF format make it practical for local deployment on modest hardware.