Meta has launched Muse Image, a new AI image generation model developed by its Superintelligence Labs that now powers image tools across Meta AI, Instagram, and WhatsApp.
The model lets users tag other Instagram accounts in prompts to incorporate their likeness into generated images, works with a reasoning AI to plan outputs, and offers features like room redesign and direct photo editing.
A Muse Video model is in development.
What happened
Meta's Superintelligence Labs has released Muse Image, an AI image generation model now powering tools across Meta AI, Instagram, and WhatsApp, with rollout to Facebook and Messenger coming soon. Users can @mention other Instagram accounts in prompts to incorporate their likeness into generated images, and the model can redesign rooms, transform images with suggested prompts, and enable direct editing by drawing on photos.
Why it matters
Muse Image replaces Meta's Llama lineup as part of a broader shift toward a new family of AI models. The model works with Muse Spark (a large language model that reasons through prompts, searches the web, and plans before generating) and positions Meta to compete in image and video generation. Users retain control over how their likeness can be reused for AI, addressing a potential concern around consent.
What to watch
Meta is planning to launch a Muse Video model, which Wang says is "competitive on prompt adherence, visual fidelity, temporal consistency." Thirty new AI effects are coming to Instagram Stories in the US before rolling out to other countries and areas of Meta's apps.
Ask the AI about this article →
Meta's launch of Muse Image marks a shift from its Llama-based approach toward a new family of AI models centered on reasoning and multimodal capability. By positioning Muse Image as "agentic"—capable of reasoning, web search, and planning—Meta is addressing a limitation of simpler generative models that simply follow instructions without intermediate reasoning. The integration with Muse Spark, a large language model, reflects a broader industry trend toward combining text understanding with image generation to improve adherence to user intent.
The ability to tag other Instagram users and incorporate their likenesses raises consumer concerns around consent and identity use, which Meta has anticipated by building in user controls for how photos can be reused. This design choice suggests Meta is aware of regulatory and reputational risks and is positioning the feature as user-friendly rather than purely exploitative. The rollout strategy—starting with Meta's existing apps and core platforms—gives Meta a built-in distribution advantage over smaller AI image startups that lack comparable user reach.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
AI system scaling has pushed interconnect requirements inside data centers from chips and boards up to racks…
Chinese large-model developer Z.ai says it can now support large-scale inference using roughly 100,000 domesti…
Analyst Ming-Chi Kuo says Nvidia has revived the Rubin CPX AI accelerator with a substantially redesigned arch…
Palantir Technologies stock has posted multi-year gains, including an 11x return over 3 years
Apple has escalated its legal battle against OpenAI, claiming in a new court filing that OpenAI is actively de…
Samsung Electronics has locked up as much as 70% of its memory production capacity under long-term supply agre…
