Meta has unveiled Muse Image, the first image-generation model developed by its new Meta Superintelligence Labs (MSL), introducing advanced artificial intelligence features that promise to make creating and editing images more intuitive across its family of apps.

The new AI model, now available through Meta AI, enables users to generate high-quality images from conversational prompts, edit existing photos using natural language instructions and create visual content directly within Meta’s ecosystem.

Muse Image will also power new creative features on Instagram and WhatsApp, with support for Facebook, Messenger and Meta Advantage creative tools for advertisers expected to roll out in the coming weeks.

Unlike conventional AI image generators that rely mainly on text prompts, Muse Image combines image generation with reasoning capabilities powered by Muse Spark, Meta’s reasoning model. According to the company, the AI plans image layouts, retrieves relevant web information when required and intelligently combines multiple visual references before creating an image.

Muse Image: AI that understands what users want

Meta says the new model is designed to make image creation feel more conversational and natural.

Users can create images from scratch or modify existing photographs simply by describing the desired changes. They can remove unwanted objects, redesign living spaces using products sourced from the web or Facebook Marketplace, generate functional QR codes and even create infographics with clearly rendered text.

Announcing the launch, Meta described Muse Image as the first image-generation model from Meta Superintelligence Labs to be integrated into Meta AI.

“The model,” Meta says, “uses advanced reasoning to understand complex prompts, seamlessly blending multiple photos into high-quality creations you can download and share anywhere, including directly to your chat, story, or feed.”

The company explains that Muse Spark works behind the scenes by planning image composition, gathering relevant real-time information where necessary and intelligently blending multiple visual references into a single result.

To make image creation easier for everyday users, Meta AI now includes a presets panel featuring suggested prompts.

The shortcuts allow users to restore old family photographs, experiment with trending hairstyles or transform images into artistic styles such as claymation or retro 16-bit video game graphics without having to write detailed prompts.

Another new capability allows users to incorporate public Instagram content into AI-generated images by tagging Instagram accounts.

“You can also @-mention Instagram accounts in the Meta AI app to bring specific Instagram profiles right into your images,” Meta says.

The company notes that users remain in control of whether their public Instagram content can be used in this way.

Muse Image also introduces visual editing tools that allow users to indicate changes by drawing directly on an image instead of repeatedly rewriting prompts.

“To make a change, just tap the markup icon on your creation to circle, sketch, or annotate edits right on top of the photo,” Meta explains.

The feature is designed to make editing faster and more precise by allowing users to highlight exactly which parts of an image should be changed.

Muse Image AI coming to Instagram Stories and WhatsApp

Beyond the standalone Meta AI experience, Muse Image will power more than 30 AI-generated Story effects on Instagram and support image generation within Meta AI chats on WhatsApp.

The WhatsApp rollout will begin in selected countries before expanding globally.

Meta says everyday image creation will remain free through Meta AI, while users who require higher usage limits and additional creative features will be able to access them through subscription plans.

Muse Video coming next

Alongside Muse Image, Meta also previewed Muse Video, an upcoming AI model designed to generate videos from user prompts.

Muse Image is the first media-generation model built by Meta Superintelligence Labs, the AI division established to accelerate the development of advanced foundation models and strengthen Meta’s position against rivals including OpenAI, Google and Anthropic.

The model builds on Muse Spark, Meta’s reasoning system introduced earlier this year to improve planning and reasoning capabilities across Meta AI.

According to Meta, Muse Image ranks among the leading image-generation systems in human preference benchmarks for both text-to-image creation and image editing. The company also says Muse Video has posted promising early benchmark results ahead of its wider release.

To promote transparency, every image created with Muse Image includes Content Seal, Meta’s invisible watermarking technology designed to identify AI-generated content even after an image has been cropped, resized or compressed.

AI moves deeper into Meta’s apps

The launch highlights Meta’s strategy of embedding generative AI directly into the apps people already use every day rather than limiting advanced AI capabilities to standalone tools.

By integrating Muse Image into messaging, social media and advertising products, Meta aims to make AI-powered image creation a seamless part of everyday communication, content creation and digital marketing across its platforms.