OpenMOSS releases MOSS-Image-SFT text-to-image model
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.
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.
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.
Model page: GPT-6 →
OpenMOSS releases MOSS-Image
huggingface.co · 11 October 2026
Loading the full article…
This text was published by huggingface.co. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Reddit discussionreddit.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Microsoft AI chief says in-house model beats OpenAI's using quarter of compute · 2 src
- Meta releases Muse Spark Open, an 816B-parameter open-weight multimodal model with 42B active parameters and 1M-token context · 24 src
- [AI News 10/16] Meta and Apple join forces. The era of running 816B massive AI on Macs is here. New announcements from NVIDIA, CERN, and Amazon|Haruki|AIを試す大学生 · 1 src
- Liquid AI releases d1-ultra decision model at $0.042 per million tokens · 1 src
- Google launches Gemini 4 Argon model globally for Pro and Ultra subscribers with introductory API pricing · 32 src
Comments
via GitHub Discussions