Alibaba Qwen releases Qwen-Image-2.1-Turbo, an image model that generates and edits in 8 steps

Alibaba Qwen/Qwen models

Alibaba's Qwen team released Qwen-Image-2.1-Turbo on 9 October 2026, an accelerated checkpoint of its 7B Qwen-Image-2.1 model. It generates and edits images in 8 denoising steps, and its weights are published on Hugging Face. Alibaba Cloud Model Studio also serves the Turbo model through its Pro and Turbo APIs.

• Runs text-to-image generation and multi-reference editing in 8 steps; the checkpoint carries its own sampling schedule, which Diffusers loads automatically.
• Uses the same 7B visual generation architecture as the base Qwen-Image-2.1.
• Weights are published as Qwen/Qwen-Image-2.1-Turbo on Hugging Face.
• The repository is under the Qwen Research License Agreement; the page does not say whether that covers the Turbo checkpoint or allows commercial use.
• The release page gives no benchmark figures.