Meta launches Muse Image and previews Muse Video

Meta announced Muse Image and previewed Muse Video in an official Meta AI post on July 7, 2026, making the launch the clearest new AI product move inside the latest 24-hour news window. Both models were developed by Meta Superintelligence Labs, the AI group Meta has been building to compete more directly with OpenAI, Google and other frontier labs.

Muse Image is now available in the Meta AI app and on meta.ai. Meta says it is also powering Instagram Stories in the United States and WhatsApp in a limited set of countries, with Facebook support coming later. Muse Video is not broadly available yet, but Meta says it is coming soon to creators and Meta AI.

The technical pitch is that Muse Image is not just a prompt-to-picture system. Meta describes it as an agentic image model that can use tools, search the web for context, write and run code for things like plots or QR codes, refine its own outputs, and spend more test-time compute when a request is difficult. It also works with Muse Spark, Meta's multimodal reasoning model, so the system can plan a visual result before generating it.

For everyday users, the practical change is distribution. AI image generation is moving from specialist tools into the social apps where people already make posts, stories, chats, invitations and edits. Meta is also adding direct photo markup, presets, room redesigns using real products, and the ability to bring public Instagram profiles into creations through @ mentions. That last feature will matter for privacy: Meta says users have controls for how their content can be tagged for AI creation, but public social media content is now closer to becoming raw material for generative tools.

For creators and businesses, the launch raises the competitive pressure around fast visual production. Meta says advertisers and agencies will get access through Advantage+ creative in the coming weeks. That could make AI-generated campaign variations more normal inside Meta's ad ecosystem.

The broader signal is simple: generative media is becoming platform infrastructure. The question is no longer only which model makes the best image. It is who controls the social graph, the distribution surface, the provenance system and the default settings around consent.