GitHub radar
An image model that keeps text inside pictures readable
MOSS-Image is a ComfyUI model that draws and edits images without turning letters into mush, and this pack already sorts every file into place.
This is a ready-made file pack of the MOSS-Image model for ComfyUI — a visual tool for running image models without writing code. It draws and edits pictures where text stays readable.
When it helps
Useful for covers, diagrams, or memes that need clean readable text inside the image. Skip it if your GPU has less than 8GB or text inside images doesn't matter to you.
Pros
- Files already sorted into ComfyUI's folders — no guessing
- Ready-made workflow templates let you start right away
- Text inside images stays readable instead of blurry
- Apache 2.0 license — free for commercial use
Cons
- Needs GPU with 8GB+, 10GB free space for full version
- The model itself isn't smarter — only setup got easier
How to set it up — step by step
- 1Open your coding AI agent inside your installed ComfyUI folder.
- 2Send the agent the text from the block below.
- 3Wait while the agent sorts the files into models/diffusion_models, loras, text_encoders, and vae.
- 4Check that a ready-made text-to-image workflow template opened.
- 5Run a cover generation with a caption and check the text in the image is readable.
Text for your agent
Copy this and send it to your agent — Claude Code, Codex, any of them:
Set up the text-in-image model from this package for me: https://huggingface.co/Comfy-Org/MOSS-Image — place the files into the ComfyUI folders, open the ready workflow template, and generate a cover with a readable caption for me.
A regular laptop with a GPU of 8 GB or more - the compressed build of the model takes about 5 GB, the full one about 10 GB.
Open on Hugging Face▌ More finds