Muse Image: Meta's New AI Image Generator

Meta has launched Muse Image, a new neural network for image generation that creates images from text descriptions and edits existing photos.

What Muse Image can do

Muse Image works in two main modes: it creates images from a text description and edits existing photos. The features are available directly inside the Meta AI chatbot, the company's AI assistant, which is already built into WhatsApp, Instagram, Facebook and Messenger.

That means users will be able to create and edit images right where they already chat and consume content, without switching to third-party services.

• Launch date: July 8, 2026
• First stage: integration into the Meta AI chatbot
• Next stage: integration into Instagram and WhatsApp
• Capabilities: creating images from text and editing existing photos
• Developer: the Superintelligence Labs division under Alexandr Wang

Why this is a strategic shift for Meta

Before Muse Image, Meta relied on its own early developments and on external solutions for image generation. Building the model inside Superintelligence Labs means the company is taking this segment under full in-house control.

Superintelligence Labs is a new Meta division formed in 2026 specifically to develop advanced AI systems. Its head, Alexandr Wang, joined Meta in early 2026. Previously he founded Scale AI, one of the leading companies in preparing data for training AI models and a strategically important player for the whole industry. His appointment as Chief AI Officer was read as a signal of a new level of ambition for Meta in the AI race.

Muse Image is the first public result of the division's work under his leadership. That makes the launch a double event: both a new product for users and confirmation that Superintelligence Labs produces real systems rather than remaining a research project.

How Meta competes with OpenAI and Google

Integrating image generation into Instagram and WhatsApp is a bet on scale. The combined monthly audience of the two platforms exceeds 3 billion active users, more than any competitor in this segment.

DALL-E is built into ChatGPT; Imagen into Gemini. Midjourney targets a creative audience and works as a standalone product. By launching Muse Image inside Instagram, Meta makes AI image generation part of everyday consumer behavior: not a separate tool, but a built-in feature in products that billions of people open every day. Neither OpenAI nor Google can reproduce that reach through their own platforms.

What this means

Meta is no longer playing catch-up in AI image generation: the first in-house model from Superintelligence Labs is a bid for an independent position alongside OpenAI and Google. The scale of Instagram and WhatsApp could make Muse Image the most widely used AI image model in the world, provided its quality proves to be on par with competitors.

Frequently asked questions

What is Meta's Superintelligence Labs?

Superintelligence Labs is an internal Meta division created in 2026 to develop advanced AI systems. It is led by Alexandr Wang, the founder of Scale AI, who took the post of the company's Chief AI Officer in early 2026. Muse Image is the lab's first public product.

When will Muse Image arrive in Instagram and WhatsApp?

As of July 8, 2026, the model is already rolling out in the Meta AI chatbot. The company has not announced exact dates for when the feature will reach all Instagram and WhatsApp users.

*Meta has been designated an extremist organization and is banned in Russia.

Where can I use Muse Image?

The model is built into the Meta AI chatbot in WhatsApp, Instagram, Facebook and Messenger, and will soon be fully rolled out in Instagram and WhatsApp.

What can you do in Muse Image?

Create images from a text description and edit existing photos.

What capabilities does the Muse Image neural network have?

Creating images from a text description and editing existing photos.

Where is the Muse Image neural network available?

In the Meta AI chatbot built into WhatsApp, Instagram, Facebook and Messenger.