Meta launches Muse Image, its first AI image model, aiming to draw creators and advertisers

Meta announced on Tuesday the release of Muse Image, its first artificial intelligence model for creating images, as the company expands efforts to keep creators engaged and to make ad production faster for brands. The new model, developed by Meta Superintelligence Labs, is intended to support image generation inside Meta’s consumer products and to provide image-generation assistance for advertiser creative workflows.

Meta said Muse Image was originally codenamed “Mango” and is the second major release from Meta Superintelligence Labs led by Alexandr Wang. In April, the unit unveiled Muse Spark, which Meta described as a large language model that succeeded its prior Llama family of models, according to the announcement covered by CNBC.

According to Meta’s rollout plan described in the coverage, Muse Image will be available for free for consumers to access through the Meta AI app and on Meta’s website, as well as through WhatsApp direct messages and Instagram Stories. The company also framed the release as a way to let users and creators generate images directly in familiar product environments rather than requiring separate third-party tools.

Meta also said Muse Image will be used for advertisers, building on its AI-powered Advantage Plus offering. In that context, Meta described using Muse Image to generate advertiser-specific image variations, aiming to produce “high-quality, on-brand ad variations with fewer iterations,” while reducing how many separate rounds may be needed to refine creative assets for campaigns.

Meta said it has been working with businesses and advertisers as it debuted Muse Image, and that, over the coming weeks, advertisers and agencies can expect to see image variants powered by Muse Image. The company’s description in the reporting emphasized native reasoning for adjusting elements, swapping styles, and creating variations based on an advertiser’s existing creative.

The move positions Meta to compete in a crowded AI image-generation market, where large tech platforms and specialized AI developers are offering tools for creating images at scale. For advertisers, the practical focus is on lowering production friction and accelerating the cycle of generating and testing ad creative, while for consumers the emphasis is on making image generation accessible inside widely used social and messaging services.

Meta’s decision to bring Muse Image directly into its own messaging and social surfaces also raises questions for users and marketers about how AI-generated content will be produced, labeled, and reused across platforms, particularly when advertiser-specific outputs are generated from existing creative. The details of how those controls and labeling mechanisms work were not included in the coverage and were not confirmed in the research materials provided for this story.