Summary

MOSS-Image is an image-model family with native layout control, typography generation, and a compact 5B backbone. It accepts ordinary text or JSON descriptions of objects, text, and bounding boxes through a shared Diffusers pipeline.

Key Features

Text-to-image generation.
JSON layout prompting.
Object bounding boxes.
Text rendering.
Up to 4K-token prompts.
Anime and photorealistic styles.
Python and Gradio interfaces.
Apache-2.0 weights.

About MOSS-Image


Developers can create structured posters and visual compositions while adapting base or fine-tuned checkpoints to their own design workflows. Core capabilities include Text-to-image generation, JSON layout prompting, Object bounding boxes, and Text rendering.


Public code or model weights are linked from the project, with setup instructions for running or adapting the system. Additional characteristics include Up to 4K-token prompts, Anime and photorealistic styles, Python and Gradio interfaces, and Apache-2.0 weights.

Get more likes & reach the top of search results by adding this button on your site!

Embed button preview - Light theme
Embed button preview - Dark theme
TurboType Banner

Subscribe to the AI Search Newsletter

Get top updates in AI to your inbox every weekend. It's free!