Meta has released Muse Image, the first AI image-generation model built by its Meta Superintelligence Labs, now available in the US via the Meta AI app, Instagram and WhatsApp. It follows Muse Spark, launched in April 2026, and marks Meta's push into image creation, matching rivals such as Google and OpenAI. The model matters because it is deeply woven into Meta's existing services and can draw on users' likenesses and product listings, signalling how the company intends to fold generative AI into its everyday apps.

Meta says the model uses advanced reasoning to interpret complex, conversational prompts, blend multiple photos, and produce legible styled text and even functional QR codes. Notable features include redecorating a room using furniture from the web and Facebook Marketplace, and incorporating a friend's likeness by tagging their Instagram account, subject to existing photo-reuse controls. Muse Image also powers over 30 new Instagram Stories effects and image editing in WhatsApp. It is free for "everyday creation", with a paid Meta One subscription for higher limits, and a Muse Video model is said to be in development.