Meta Muse Image

What Are Meta Muse Image AI Prompts?

Meta Muse Image is the image generation model Meta Superintelligence Labs released on July 7, 2026. Instead of mapping a prompt straight to pixels, it plans the layout, looks up the context it needs on the web, and refines its own output before returning it. Carat pairs each Muse Image result with the prompt behind it, so you can copy and run it as is.

Where You'll Use It

• Posters and exploded-view diagrams where text, tables, charts and labels have to be correct
• Composites that pull the person from one reference, the background from another, and props from a third
• Restyling a single photo into oil paint, clay, pixel art or papercraft
• Interior restyling that keeps the room's layout and swaps only the furniture and decor
• Product cuts relit with studio lighting and a new backdrop for a detail page
• Restoring old photos by clearing scratches and correcting color

Why This Tag Is Useful

Muse Image is tuned to follow written instructions in the order you wrote them, so conversational sentences work better than comma-separated keyword strings. A single edit request takes up to 10 reference images, and follow-up requests keep the person and framing from the image you just made. Rendering fine detail such as text and charts accurately is one of the axes the model was built around. You can pick from nine aspect ratios between 21:9 and 9:21, and if you leave it out the model sets the dimensions from the request.

Tips for Crafting Prompts

1. Write what changes and what stays as separate sentences to limit unintended drift.
2. Put any text that belongs in the image inside quotes, and say where it should sit.
3. When you supply several references, state what each one is for: the subject, the background, the props.
4. Set the aspect ratio in the prompt or the settings so the output fits where it is going.

When the result misses your intent, adding one sentence about the part that is off is faster than rewriting the whole prompt.