The Chroma Workflow combines the power of the 8.9B parameter Chroma model, built on Flux.1 Schnell, with a suite of tools designed to elevate your image generation projects. Still under active development, Chroma is Apache 2.0 licensed, meaning it’s open-source and fully customizable. Whether you’re working with text-to-image (txt2img), image manipulation, or upscaling, this workflow has everything you need to bring your creative visions to life.

🔹 Workflow Features:

• txt2img & img2img – Generate new images from text prompts or modify existing ones.
• Detail-Daemon – Add intricate details to your images for a more polished and realistic result.
• Inpaint – Seamlessly fill in missing areas of images for better composition.
• HiRes-Fix – Enhance image resolution and overall quality, bringing out fine details.
• Ultimate SD Upscale – Upscale your images while maintaining high quality and clarity.
• FaceDetailer – Improve and refine facial details for more lifelike renderings.

The Chroma model itself is still being trained, with frequent updates (latest v29.5 as of May 15th), and comes from a dataset of 5M curated samples, including anime, furry art, artistic imagery, and photography. Fully uncensored, it reintroduces anatomical details often missing in other models, making it a versatile and reliable tool for anyone in the creative community.

🔹 Original Author: Workflow

🔹 CivitAI Link: Chroma on CivitAI

🔹 Model Versions: Chroma Versions on Hugging Face

Start working with the Chroma Workflow today and unlock the potential of this powerful generative model!