Composition before pixels
Complex prompts work better when the model can interpret relationships first: what sits in front, what stays unchanged, where text belongs, and which object must remain the focal point.
Reasoning-led visual workspace
Muse Image AI Generator
Turn a detailed brief or reference image into a polished visual. Compose the scene, direct the mood, choose the frame, and iterate from one focused creative workspace.
Prompt to imageReference editingResearch updated July 23, 2026
Interactive Muse Image generator
Muse Image
Creative artboard
Studio
197 / 5000
Workspace
Muse Image
Mode
Studio
Credits
10
Sign in to generate. Your brief stays editable while the result renders.
Artboard ready
1K output
Your result will appear on this artboard.
Start with a complete visual brief, or switch to Edit and add reference images.
Model overview
What is Muse Image?
Introduced
July 2026
Creator
Meta Superintelligence Labs
Core idea
Plan, compose, refine
Muse Image is Meta's first image generation model from Meta Superintelligence Labs. Announced in July 2026, it was introduced as the image-making counterpart to Muse Spark, with a workflow that interprets a request, plans the composition, and then produces the visual.
The key idea is a creative process that goes beyond a one-line prompt. Meta describes Muse Image as able to combine visual references, use current web context, render structured elements, and support direct edits on an image. That makes it relevant to creators who need a campaign draft, room concept, social asset, product scene, or visual explanation rather than an isolated piece of AI art.
This page gives you a focused browser workflow for prompt-to-image creation and reference-led editing. Write a complete brief, add your own source images when identity or composition matters, choose the final frame, and evaluate the result as a publishable asset.
Creative strengths
Why Muse Image changes the prompt-to-picture workflow
Muse Image is positioned as a creative partner that can reason about the whole deliverable. These are the practical ideas worth carrying into your own brief.
Reference-aware creation
Multiple references can separate identity, product shape, palette, texture, and environment. A precise brief tells the model what to borrow from each source instead of blending everything indiscriminately.
Structured visual communication
Meta highlights cleaner in-image text, infographics, guides, and even functional QR concepts. For production work, keep the required copy short and inspect every character before publishing.
Iterative image editing
Muse Image was introduced with markup-led edits and conversational refinement. In this workspace, switch to Edit, upload references, and describe only the change you want while naming the details to preserve.
Original case studies
Muse Image examples for real creative work
Each example starts from a commercial outcome, not a style label. Use the prompt notes to make your own request more specific.
Case 01Fragrance launch campaign
A glass perfume bottle becomes the hero of a warm, editorial product scene with controlled reflections and generous copy space.
Prompt note
Define the product material, surface, light direction, palette, lens feel, and where the campaign headline will sit.
Case 02Room redesign concept
A compact living room is restyled with warm wood, sculptural seating, soft daylight, and believable product placement.
Prompt note
Name what must stay fixed in the room, then list the materials, furniture mood, lighting, and shopping constraints.
Case 03Editorial character portrait
A distinctive subject remains consistent while wardrobe, color, and art direction create a magazine-ready portrait.
Prompt note
Separate identity details from styling instructions and specify crop, gaze, background, wardrobe, and finish.
Case 04Product story infographic
A clean visual guide turns a reusable bottle into a structured product explanation with clear hierarchy and material callouts.
Prompt note
Keep labels short, state the exact words, define the reading order, and request no additional text.
Practical workflow
How to use the Muse Image generator
Treat the first prompt like an art-direction brief. A deliberate first pass gives you a stronger image and makes later edits easier to control.
- 01
Choose Create or Edit
Use Create for a new scene. Use Edit when a product, person, room, layout, or visual identity must come from an existing image.
- 02
Write the visual hierarchy
Start with the main subject and final use. Add composition, environment, materials, lighting, camera, palette, and anything the model must avoid.
- 03
Set the publishing frame
Choose 1:1 for flexible social and catalog crops, 16:9 for wide campaigns, 9:16 for stories, or a portrait ratio for editorial work.
- 04
Inspect, then refine
Check anatomy, product identity, text, reflections, and small objects at full size. On the next pass, ask for one controlled change and restate what must remain untouched.
Prompt craft
A Muse Image prompt formula that stays controllable
Good prompts describe decisions. They do not need decorative adjectives in every sentence.
01Deliverable02Primary subject03Composition04Environment05Materials06Lighting07Camera08Color09Exact text10Preserve / avoid
Example brief
Create a 4:5 editorial launch image for a refillable citrus fragrance. Center a translucent amber bottle on pale limestone, leave clean space in the upper third, use low late-afternoon sunlight from camera left, preserve realistic glass thickness and a soft grounded shadow, warm cream and electric-yellow palette, 85mm product photography, no logo, no readable text, no extra bottles.
Where it fits
Best uses for Muse Image
Campaign concepts
Explore hero images, art direction, backgrounds, and crop variations before a full production shoot.
Ecommerce scenes
Place a product in styled environments while preserving its form, finish, and recognizable details.
Social storytelling
Create feed, story, and chat-friendly visuals with intentional framing and reusable visual motifs.
Interior ideation
Restyle an existing room photo while keeping the architecture and spatial proportions understandable.
Portrait art direction
Use a reference to explore wardrobe, editorial lighting, illustration, or period-inspired treatments.
Visual explainers
Draft product callouts, guides, diagrams, and infographics, then verify text and facts before release.
Decision guide
Muse Image versus a one-shot image workflow
The real difference is how you structure the work, not how many adjectives you place in a prompt.
| Workflow | Muse Image approach | Basic one-shot approach | Best choice when |
|---|---|---|---|
| Complex composition | Plan relationships and layout before rendering | Translate one prompt directly into a scene | Several objects or spatial rules must align |
| Reference images | Assign separate roles to identity, style, and setting | Use one reference as loose visual guidance | A product or character must stay recognizable |
| Editing | Refine a target while preserving named details | Regenerate the full image and compare | You need controlled iteration rather than a restart |
| Structured content | Plan hierarchy for text, guides, and layouts | Treat text as another visual texture | The image must communicate, not only decorate |
Availability and trust
What to know about Muse Image access in 2026
Meta launched Muse Image inside Meta AI and announced related experiences for Instagram and WhatsApp, with more surfaces planned. At launch, Meta had not published a standalone public developer API or independent rate card for the model.
The launch also included a way to reference public Instagram accounts. After criticism about consent and likeness use, Meta updated its announcement on July 10, 2026 and removed that specific feature. The broader lesson is useful for every image workflow: only upload material you have the right to use, get permission before recreating a real person's likeness, and review generated visuals before sharing them.
Model features and availability can change quickly. This guide links to the dated primary announcement and its update so you can distinguish current facts from early launch claims.
Questions answered
Muse Image FAQ
What is Muse Image?
Muse Image is the first image generation model announced by Meta Superintelligence Labs. Meta introduced it in July 2026 as a reasoning-led creative model that works with Muse Spark to plan layouts, combine visual references, and support image generation and editing.
Can I use Muse Image from a text prompt?
Yes. Start in Create mode, describe the final deliverable and main subject, then add composition, lighting, materials, camera, color, and constraints. A structured brief is more controllable than a list of style adjectives.
Can Muse Image edit an existing photo?
Muse Image was announced with photo editing and markup-led refinement. In this workspace, choose Edit, upload your own reference image, describe the change, and state which identity, product, framing, or background details must remain fixed.
Does Muse Image support multiple reference images?
Multi-reference composition is one of the capabilities highlighted in Meta's launch material. When using references, assign each image a job, such as product identity, color palette, wardrobe, or environment, so the model does not merge them without direction.
Is Muse Image good at text inside images?
Meta emphasizes cleaner text rendering for guides and infographics, but every generated word still needs human review. Keep copy short, specify the exact text, forbid extra wording, and check small labels at full resolution.
Is there a public Muse Image API?
At the time of this July 2026 update, Meta had announced Muse Image across its own products but had not published a standalone public developer API or public API pricing. Availability may change, so verify against Meta's current developer information.
How should I use portraits and public photos responsibly?
Use images you own or have permission to use. Get consent before generating a recognizable real person, avoid deceptive or harmful contexts, and review the final result. Meta removed the launch feature that referenced public Instagram accounts after consent concerns.
Start with a brief worth rendering
Describe the deliverable, subject, layout, light, and constraints. Muse Image works best when the creative decision is clear before generation begins.