Meta launches Muse Image AI model for image generation and editing

Meta has unveiled its latest AI image model, Muse Image, developed by Meta Superintelligence Labs. The model is now available to users in the United States through the Meta AI app, Instagram, and WhatsApp.

Muse Image uses advanced reasoning to understand complex prompts and seamlessly blend multiple photos into high-quality creations. It can generate images in a variety of formats and styles, and it understands conversational prompts, making it easy to request changes. Meta claims the model can generate functional QR codes and produce legible, stylized text within images.

The new model powers over 30 new effects for Instagram Stories, including one that edits a photo to look like it was taken with a disposable camera and another called "Puffer." A new interface in Instagram's image editor lets users preview these effects before applying them, and they can also prompt an entirely new edit if they are not satisfied with the built-in effects.

Like Google, Meta is connecting its AI model to other services. For instance, taking a picture of a room and asking Meta AI to redecorate can pull furniture listings from the web and Facebook Marketplace. Tagging a friend's Instagram account can incorporate their likeness into a generated image. The same controls users have in place for how their photos are reused on Instagram apply.

Using Muse Image is free for everyday creation, but there are limits. Meta said limits could vary depending on variables such as which service users are accessing from or their current location. Once the limit is reached, Meta prompts users to subscribe to Meta One before their free image allotment resets. The model is available in the US to start, with more countries and products to follow. Additionally, Meta confirmed that a Muse Video model is in active development.