Meta is letting anyone use its Muse model for free — no API keys, no code required

Meta is opening up Muse, its image-generation model, to the public. Anyone can use it today at muse.meta.com without an account. Upload an image, add a prompt, and the model re-renders it — resizing, restyling, or regenerating parts of the image entirely.

The model is built on Meta's Llama 4. It's been trained on a curated dataset of high-quality images and is designed to handle 20+ different image-generation tasks without needing separate fine-tuned models for each one. Meta released the weights alongside the tool, and the open-source code is on GitHub.

This is not a new model being wrapped in a new interface. It's the same architecture Meta uses internally, now available as a free tool. The weights are available for anyone who wants to run their own instance — no vendor lock-in, no API rate limits.

Muse is one of the more useful image models to come out of a big lab in a while. It doesn't try to do everything; it does a focused set of tasks well. And because it's open-weight, you're not paying Meta to use it — you can run it on your own hardware or in any cloud that supports the weights.

Why this matters for us: a free, open image model means any community org, small shop, or tía with a Facebook page can generate clean, usable images without paying for another SaaS subscription.