# OpenMOSS releases MOSS-Image-SFT text-to-image model

Digest AI · Generative AI & Models · published 2026-10-11T00:00:00Z

Canonical: https://digestai.news/story/openmoss-releases-moss-image-sft-text-to-image-model

## Summary

OpenMOSS has released MOSS-Image-SFT, a supervised fine-tuned text-to-image model derived from the MOSS-Image-Base architecture. The release supports natural-language prompts, tag sequences, and JSON layouts to control text placement and object arrangement within generated images. The model is designed to produce clearer text grouping and more organized layouts compared to its base version.

The model requires Python 3.10 or later and an NVIDIA GPU with CUDA support. It utilizes PyTorch 2.9 or later and specific versions of libraries such as diffusers, transformers, and accelerate. The transformer and text encoder operate in bfloat16 precision, while the VAE and initial noise use float32. The bundled VAE uses the FLUX.2 small decoder.

OpenMOSS provides a Python API for generation, where image dimensions must be positive multiples of 16. The release also includes a Prompt Enhancer tool designed for GPT-6-family agents, which converts image ideas into detailed captions or spatial Layout prompts. The model weights and inference code are released under the Apache License, Version 2.0, with integration into ComfyUI currently in progress.

## Key points

- MOSS-Image-SFT is a fine-tuned text-to-image model from OpenMOSS derived from MOSS-Image-Base.
- The model supports JSON layouts for controlling text placement and object arrangement in images.
- Weights and code are released under the Apache License, Version 2.0, requiring NVIDIA GPUs.

## Why it matters

The release provides an open-weights option for structured image generation, allowing developers to control text and object placement via JSON layouts, which is useful for creating diagrams and infographics.

## Sources

1. [OpenMOSS releases MOSS-Image](https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT) (huggingface.co, 2026-10-11, primary source)

## Cite

Digest AI, "OpenMOSS releases MOSS-Image-SFT text-to-image model", 11 October 2026, https://digestai.news/story/openmoss-releases-moss-image-sft-text-to-image-model

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/openmoss-releases-moss-image-sft-text-to-image-model.json