Meta Muse Image is Meta's new image generation and editing model inside Meta AI. It can create images from text, edit an uploaded photo, combine several references, and refine the result across a conversation.

The interesting part is not another "type a prompt, get a picture" workflow. Meta says Muse Image can plan a composition, use search or code for some tasks, inspect its own draft, and make corrections before returning the result.

This is an early guide based on Meta's launch materials and current product availability, not a claim that we independently reproduced every benchmark. Features can also vary by country and account while the rollout continues.

What Is Meta Muse Image?

Muse Image is the first image generation model from Meta Superintelligence Labs. It works with Muse Spark inside Meta AI and is designed for both generation and editing.

According to Meta's technical announcement, the model can use search, generate code for precise graphics such as plots and QR codes, refine its own drafts, and compose people, objects, clothing, styles, and environments from multiple reference images.

Text-to-image

Describe a scene in conversational language and continue refining it in chat.

Image editing

Upload a photo, describe the change, or mark the area that needs attention.

Multi-reference composition

Combine authorized reference images for a person, object, wardrobe, style, or location.

Text and structured graphics

Create layouts that include labels, signs, invitations, diagrams, or other visible text.

Where Can You Use Muse Image?

As of July 25, 2026, Meta says Muse Image is available in the Meta AI app and on meta.ai. It is also available in Instagram Stories in the United States and in WhatsApp in limited countries. Facebook, Messenger, more regions, and advertiser tools are still rolling out.

If the option is missing, that does not necessarily mean your account is broken. Check the Meta AI app or website first, then confirm whether the feature has reached your region.

Important July 10 update

Meta removed the feature that let people reference public Instagram accounts with an @-mention. The official launch post now confirms it is no longer available. Use photos you own or have permission to edit.

A Better Prompt Formula for Muse Image

Start with the result you need, not a pile of style words. This five-part structure gives the model a clear job:

Create or edit: [the exact deliverable] Subject and setting: [who or what, where, and what is happening] Composition: [camera angle, crop, focal point, and empty space] Visual direction: [lighting, color, texture, and mood] Constraints: [text to include, details to preserve, and what not to change]

For edits, state what must stay unchanged. For multi-reference work, explain the role of each image instead of asking the model to guess.

12 Meta Muse Image Prompts to Try

1. Natural editorial portrait

Create a natural editorial portrait of a ceramic artist in a bright working studio. Waist-up composition, hands lightly dusted with clay, soft window light from camera-left, realistic skin texture, muted earth tones, 50mm portrait look, quiet and confident mood. Keep the background recognizable but gently out of focus. No beauty-filter smoothing.

2. Product launch image

Create a square product launch image for a matte forest-green insulated bottle. Place the bottle on pale stone with a soft shadow and a small pool of water nearby. Clean studio lighting, crisp material texture, premium outdoor brand mood, empty space in the upper-left for a headline. Do not add a logo or invented text.

3. Instagram Story announcement

Design a 9:16 Instagram Story for a neighborhood coffee shop's summer cold brew launch. Warm cream background, deep brown type, one realistic glass of cold brew with condensation, subtle hand-drawn sun accents. Include only this text: "Cold Brew Season" and "Available Friday." Keep every word large and legible.

4. Room redesign from a photo

Using my room photo, redesign the space as a warm modern reading room. Keep the windows, floor, doors, and room dimensions unchanged. Replace only the furniture, rug, lamps, and wall color. Use walnut, oatmeal linen, olive accents, and soft evening light. Preserve realistic scale and walking space.

5. Remove a distracting object

Remove the trash bin behind the subject. Reconstruct the brick wall, pavement, and shadow naturally. Do not change the person's face, hair, clothing, pose, body shape, color grade, crop, or depth of field.

6. Combine three authorized references

Use reference 1 for the person's face and hair, reference 2 for the red jacket, and reference 3 for the rainy Tokyo street setting. Create a waist-up nighttime portrait with realistic reflections and soft neon light. Preserve the face from reference 1 and the exact jacket design from reference 2. Do not blend identities.

7. Brand mascot variation

Using the uploaded mascot sheet, create the same fox mascot packing an online order at a small studio desk. Preserve the face shape, orange and cream color pattern, green apron, eye style, and flat vector treatment. Change only the pose, props, and setting. Simple background, clear silhouette, no extra limbs.

8. Food photo for a menu

Create a realistic menu photo of mushroom ravioli in a shallow white bowl. Three-quarter camera angle, soft restaurant window light, visible pasta texture, light parmesan, small sage leaves, natural sauce sheen, dark walnut table, restrained styling. Keep the full plate inside the frame and avoid exaggerated steam.

9. Simple how-to graphic

Create a clean vertical how-to graphic titled "Three Steps to Better Product Photos." Show three numbered panels: 1. Place the product near a window. 2. Use white card as fill light. 3. Lock exposure before shooting. Minimal cream and green palette, clear icons, generous spacing, and large legible text. Include no other words.

10. Event invitation

Design a playful 5:7 birthday invitation with watercolor dinosaurs, balloons, and confetti. Include exactly: "Maya Turns 8", "Saturday, August 15 at 2 PM", "Riverside Park", and "RSVP by August 8." Keep the hierarchy clear and verify every letter before finishing.

11. Consistent ad variation

Create a second ad variation from the uploaded product image. Preserve the product shape, label, logo, colors, and proportions exactly. Change the background to a bright bathroom shelf with morning window light. Leave clean space on the right for copy. Do not invent packaging details.

12. Custom postcard

Combine my authorized portrait and vacation landscape into a realistic travel postcard. Place the person naturally at the overlook, matching the direction and softness of the landscape light. Keep facial identity and clothing unchanged. Add a small handwritten-style caption that reads exactly "Wish you were here." Landscape 3:2 composition.

What Muse Image Still Does Not Solve

Reasoning and self-refinement can improve an output, but they do not remove the need to review it. Check hands, tiny objects, brand assets, exact text, shadows, reflections, and identity before publishing.

  • Availability is uneven. A tutorial may describe a feature your region does not have yet.
  • Benchmarks are not your workflow. Meta's launch rankings are useful context, not a promise for every prompt.
  • Text still needs proofreading. Ask for exact copy and inspect every word.
  • References require permission. Do not upload or remix someone else's likeness or work without authorization.
  • One model is not always best. Test the same brief across models when realism, editing precision, or style consistency matters.

How QuestStudio Helps

Muse Image is useful when you want a conversational workflow inside Meta's apps. QuestStudio is useful when the job is broader: organize reusable prompts in Prompt Lab, generate and compare image options in Image Lab, then keep the workflow connected to your other creative tools.

Save the five-part formula above as a reusable prompt, swap only the subject and constraints, and compare how different models interpret the same brief. That gives you a repeatable process instead of a one-off result.

Meta Muse Image FAQ

What is Meta Muse Image?

Muse Image is Meta's image generation and editing model inside Meta AI. It creates from text, edits uploaded photos, combines multiple references, and supports iterative refinement.

Where can I use Muse Image?

Meta lists the Meta AI app and meta.ai, Instagram Stories in the United States, and WhatsApp in limited countries. The rollout is expanding.

Is Meta Muse Image free?

Meta says everyday creation is free and additional use is available through subscription plans. Limits may vary by account and region.

Can Muse Image edit existing photos?

Yes. It supports conversational editing, markup instructions, and multiple reference images. Use images you own or have permission to edit.

Can Muse Image reference public Instagram accounts?

No. Meta removed that feature on July 10, 2026 after criticism. Current workflows should use authorized images that you upload yourself.

Related guides