Muse Image: Meta's new AI image model launches with @mention

Meta is launching the first AI image generation model made by its Superintelligence Labs division. Muse Image now powers the image-making tools across the Meta AI app, Instagram, and WhatsApp, and is coming soon to Facebook and Messenger.

What is Muse Image and where will it appear?

Muse Image is the latest entry in Meta’s Muse family of models, intended to replace the company’s Llama lineup, and it is already active across the Meta AI app, Instagram, and WhatsApp. Meta announced the model on July 7, 2026, and said it will roll Muse Image into Facebook and Messenger in the near future.

The Muse family is being positioned as a successor to Llama. Meta also called Muse Image part of a broader stack that includes a Muse Spark large language model; Alexandr Wang, who Meta hired to lead Superintelligence Labs, described Muse Image as “agentic,” saying it works with Muse Spark "to reason through your prompt, search the web, and plan before it generates." Meta has also teased a Muse Video model.

How does Muse Image incorporate other users into generated photos?

Muse Image lets people @ mention Instagram accounts inside prompts, enabling the model to incorporate those users’ likenesses by using public photos tied to that username. Meta explains that tagging a username lets Meta AI use public photos to build a visual, and it says users will have controls over how people reuse their content for AI.

Beyond pulling in likenesses, Muse Image supports transforming existing images with suggested prompts, redesigning rooms from images (including listings pulled from Facebook Marketplace or elsewhere on the web), and editing photos by drawing directly on top of them. Outputs can be shared to a feed, story, or chat. Meta will also use Muse Image to power 30 new AI effects coming to Instagram Stories in the US before those effects roll out to more countries and more areas of its apps.

Why it matters

Meta has embedded Muse Image across multiple consumer endpoints, which takes AI image generation from experimental labs into everyday social features on Instagram, WhatsApp, and the Meta AI app. Allowing @mentions to pull public photos into generated images changes the practical scope of image generation: it ties generation to social identities and existing public content rather than limiting outputs to abstract or generic subjects. That raises questions about content control, reuse settings, and how public media tied to accounts will be treated by systems trained or prompted to assemble likenesses.

The company’s decision to describe Muse Image as working in tandem with an LLM that can "reason," search the web, and plan before generating suggests Meta is aiming for more context-aware outputs than single-step image generators. If Muse Video ships with the same prompt-adherence and temporal consistency goals Meta hinted at, the company will be pushing toward richer, multimodal consumer features tied directly into its social graphs.

What to watch

Track the rollout to Facebook and Messenger and the availability of the 30 Instagram Stories AI effects in the US; those are the concrete product milestones Meta named. Also watch for the launch of the Muse Video model, which Meta teased as competitive on prompt adherence, visual fidelity, and temporal consistency, and for details on the user controls Meta provides for reuse of public content and likenesses.