Turn Text or JSON Layouts into Images with MOSS-Image
freeTrialImage.bannerPity

Report

Result
Preview of your layout

freeTrialImage.bannerPity

freeTrialImage.upgradeUnlock

  • ✓freeTrialImage.benefitHd
  • ✓freeTrialImage.benefitWatermark
  • ✓freeTrialImage.benefitUnlimited

Comfyui Moss Image

ComfyUI users can run Comfyui Moss Image for box-driven JSON layouts, Base/SFT checkpoints, and readable text up to 4K tokens.

AI Tools

Explore AI tools

placeholder hero

Comfyui Moss Image: Layout Precision for ComfyUI Workflows

This lean 5B image family pairs structured JSON placement with reliable text, plus anime and photoreal looks for diverse scenes.

  • Structured JSON Placement
    Use box coordinates inside a JSON prompt to define where objects, captions, and scene parts sit and how much space they take.
  • Long Prompts on a 5B Core
    Prompts can reach 4K tokens on a compact 5B model; Base and SFT variants use the same sampling flow.
  • Readable Text, Anime to Photoreal
    Clear on-image lettering meets flexible anime and photo-real aesthetics for illustrations, photos, and design layouts.

Run Comfyui Moss Image: Three Simple Steps

Go from plugin setup to a finished layout render in three steps.

FAQ

Comfyui Moss Image: Common Questions

Answers to frequent questions about MOSS-Image layouts inside ComfyUI.

1

Do I need the official ComfyUI repository?

No. Search MOSS-Image in ComfyUI Manager to install the native plugin; official repository integration is still underway.

2

How is MOSS-Image-Base different from SFT?

They share the same architecture and sampling flow; the sample shows SFT producing tighter text grouping and a cleaner whiteboard arrangement.

3

What role do bounding boxes play in a layout prompt?

Every box defines an element's location and scale; edit the text, descriptions, and coordinates to rearrange content and composition.

4

What prompt length and styles can it handle?

It accepts prompts up to 4K tokens and supports anime and photoreal looks across text, illustration, photography, and visual design.

5

Which memory modes suit low-VRAM systems?

Advanced settings include GPU resident, Component, and Sequential modes; use Sequential or smaller images when memory is limited, and note Compile DiT is off by default and not available in Sequential.

6

Are the weights fine-tunable under Apache-2.0?

Yes. The 5B backbone can be fine-tuned for custom tasks and domains; weights are Apache-2.0 licensed, while pipeline code and third-party parts keep separate licenses.

Put Comfyui Moss Image to Work in ComfyUI

Add the plugin through ComfyUI Manager, load the sample workflow, and test a JSON layout with Base or SFT weights.