A simple MOSS-Image workflow for ComfyUI. Type the exact words you want, press Run, and the model draws them inside the picture, spelled right.
Good for posters, signs, menus and covers where the words have to be correct.
Watch the video: https://www.youtube.com/watch?v=OoEZh5fmK_Y
What you need
- A MOSS picture model in models/diffusion_models. Ours are free here on CivitAI: Realism by Stable Yogi Moss (photo look) or Muse by Stable Yogi Moss (glamour). Both make a finished picture in 8 steps at CFG 1.
- A MOSS-Image text encoder, for example qwen3.5_9b_w4a8_gptq.safetensors in models/text_encoders (Apache-2.0)
- flux2-vae_small.safetensors in models/vae
- ComfyUI 0.43 or newer (older versions do not know the moss_image text encoder type)
How to use it
- Drag the workflow file onto the ComfyUI canvas.
- In the text encoder loader, set the type to moss_image and pick your three files.
- Put the exact words in quotes and say where they go, for example: a poster that says "OPEN" at the top.
- Start with 25 steps, CFG 2.5, Euler + Simple, 1024 x 1024 (with Realism or Muse Moss: 8 steps, CFG 1, ER SDE + Beta). Not happy? Change the seed and run again.
The step-by-step guide, the video and more free workflows are on my portal.

