Meta introduces Muse Image, bringing AI image generation deeper into Instagram and WhatsApp

Meta has unveiled Muse Image, its first image-generation model developed by Meta Superintelligence Labs (MSL), alongside a preview of Muse Video, marking the social media giant's biggest push yet into AI-powered creative tools.

The company says Muse Image is built to go beyond simply generating images from text. Instead, it is designed as "a creative partner that knows your world," capable of understanding complex prompts, editing images with precision, combining multiple reference photos, and even drawing context from Instagram to create more personalised outputs. "Muse Image follows instructions faithfully, edits with precision, composes from multiple references, and draws on Instagram for social context," Meta said while announcing the model. "Muse Video delivers exceptional visual fidelity with native audio support."

Unlike earlier image generators, users can upload several photos, sketch directly on images to make edits, or simply describe changes in natural language. Meta is also introducing prompt presets to help users get started and allows people to tag Instagram accounts using "@mentions" to incorporate photos into AI-generated creations.

One of the headline features is its deep integration with Meta's ecosystem. Muse Image is rolling out through Meta AI and will power more than 30 new AI effects for Instagram Stories. Users can create images directly inside chats and share them instantly across WhatsApp, Instagram, Facebook and Messenger, reinforcing Meta's strategy of embedding AI into products people already use rather than offering standalone creative tools.

Meta says Muse Image is the first media-generation model built by Meta Superintelligence Labs, the AI division created after the company reorganised its artificial intelligence efforts. The launch follows Muse Spark, the reasoning model introduced earlier this year that now powers Meta AI. Together, the Muse family represents a shift away from Meta's Llama branding for its consumer AI products.

The company also offered an early look at Muse Video, an AI model that can generate videos from text and image prompts. While Meta has not disclosed a launch timeline, it said the model is designed to produce videos with "strong prompt adherence, exceptional visual fidelity and temporal consistency," along with native audio generation.

The rollout comes as Meta races to close the gap with rivals including OpenAI and Google in multimodal AI. According to independent benchmark rankings cited by industry publications, Muse Image performs competitively against Google's latest image models, though OpenAI's GPT Image remains ahead in several evaluations.

The launch also signals Meta's long-term ambition to make AI creation a native feature across its family of apps. As the company put it, the goal is to make it possible for people to "turn your ideas into high-quality visuals that you can download and share anywhere"—without ever leaving the Meta ecosystem.