Google Research introduced Imagen as a text-to-image diffusion model in this research paper.
Muse: Text-To-Image Generation via Masked Generative Transformers
Muse is a text-to-image model that also supports editing images through its generative transformer approach.
About Muse: Text-To-Image Generation via Masked Generative Transformers
Muse is a text-to-image Transformer model presented on a research website. It generates images from text prompts and also supports editing through masked generative transformer methods, including inpainting, outpainting, and mask-free editing. The page describes training on masked image tokens conditioned on T5 text embeddings, with base and superres transformers. It reports 1.3s generation for 512x512 and 0.5s for 256x256 on TPUv4, plus 760ms updates for interactive editing.
The site focuses on model details, benchmark results, and examples. Pricing is unknown, the model is not open source, and no repository is listed.
Key features
- Generates images from text prompts using masked generative transformers.
- Enables inpainting, outpainting, and mask-free editing without fine-tuning or inversion.
- Supports mask-free editing by iteratively resampling image tokens conditioned on text.
- Provides real-time interactive editing with updates processed in 760ms.
- Uses T5 text embeddings with base and superres transformers.
- Reports inference of 1.3s at 512x512 and 0.5s at 256x256 on TPUv4.
Best for: Researchers and developers studying efficient text-to-image generation and zero-shot image editing models.
Good for
- Reports significantly faster inference than diffusion and autoregressive models in its comparisons.
- Enables inpainting, outpainting, and mask-free editing without fine-tuning or inversion.
- Reports FID 6.06 on CC3M and FID 7.88 plus CLIP 0.32 on zero-shot COCO.
Watch out for
- Not open source; no repository is listed.
- Pricing is unknown; no pricing information is provided.
- Reported inference times use TPUv4 hardware.
Muse: Text-To-Image Generation via Masked Generative Transformers at a glance
- Website
- muse-model.github.io
- Category
- AI & Machine Learning › Models & ML Research
- Type
- Website
- Pricing
- Not confirmed. Check the site
- Platforms
- Web
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
Muse: Text-To-Image Generation via Masked Generative Transformers alternatives
Tools and sites you could use instead, ranked by how closely they match.
The ERNIE blog announces model releases and shares information about training and evaluation results.
TokenVerse presents research on personalizing image generation through multi-concept token modulation.
Image Arena ranks text-to-image generators and lets users compare sample outputs side by side.
Falcon LLM is described as a generative language model from the Technology Innovation Institute for a range of AI applications.
Stable Diffusion v1.4 is a text-to-image model documented with its model card on Hugging Face.
Popular in Models & ML Research
Well-known listings from the same topic.
wTAObittensor.comAlpacacrfm.stanford.edu/2023/03/13/alpaca.htmlSmolVLAhuggingface.co/blog/smolvlaAI Indexhai.stanford.edu/ai-index
Muse: Text-To-Image Generation via Masked Generative Transformers FAQ
What tasks does Muse support?
Muse generates images from text prompts and supports inpainting, outpainting, mask-free editing, and interactive editing.
How fast is Muse?
The page reports 1.3s for 512x512 and 0.5s for 256x256 on TPUv4, with interactive editing updates processed in 760ms.
Does Muse require fine-tuning for editing?
No. It reports zero-shot inpainting, outpainting, and mask-free editing without fine-tuning or model inversion.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.