Swift-Qwen3.8-27B

Open-source post-trained Qwen 3.8 27B model that cuts overthinking, with shorter reasoning and faster output. Offers GGUF versions and a free research API.

Visit website(opens in a new tab)