MOSS-Image
Official repository for MOSS-Image, a 5B-parameter base DiT model with high-quality text rendering, layout generation, and efficient fine-tuning.
- controllable-generation
- foundation-models
- image-generation
- text-rendering
- text-to-image
- Stars
- 208
- Forks
- 9
- + today
- +1
- Created
- 6d
Ranking data as of October 21, 2026 (UTC).
Star History
Today, hour by hour
Overview
Official repository for MOSS-Image, a 5B-parameter base DiT model with high-quality text rendering, layout generation, and efficient fine-tuning. It ranks #3560 on GitTiger, gaining +1 star on October 21, 2026 (UTC).
The project is written in Python and has 9 forks. It was created 6d ago.
What it is
MOSS-Image is a Python project under the Apache-2.0 license, per the supplied metadata. Its description calls it the official repository for MOSS-Image, a 5-billion-parameter base diffusion transformer (DiT) for image generation, with text rendering, layout generation and efficient fine-tuning.
Capabilities
The README preview states: - Layout control over the placement of text and visual elements. - Text rendering aimed at clear, accurate words within images. - Prompts of up to 4K tokens. - Anime and photorealistic styles.
The preview names two variants, MOSS-Image-Base and MOSS-Image-SFT, and says code and model weights are publicly available.
Getting started
No installation steps were supplied. The preview links English and Chinese user guides, whose text was not included. See those guides or the repository README.
git clone https://github.com/OpenMOSS/MOSS-Image.git
cd MOSS-Image
# see README for setupFrequently Asked Questions
How large is the model?
The description says it is a 5-billion-parameter base DiT model.
Are the model weights public?
The README preview says code and model weights are publicly available.
What is the maximum prompt length?
The preview says prompts of up to 4K tokens are accepted.