Qwen has added Qwen-Image-2.1‑Turbo to Hugging Face. It is an accelerated checkpoint of the 7 billion‑parameter Qwen‑Image‑2.1 model for text‑to‑image generation and image editing. The model runs with a built‑in 8‑step denoising schedule, so users do not need to configure a sampler. It loads directly with the QwenImage21Pipeline in the Diffusers library and supports several aspect‑ratio presets up to 2048 × 2048 pixels.
Why it matters
Developers can generate high‑resolution images with fewer steps, cutting compute time for creative projects.
