Specifications
- Input
- Text prompt or uploaded photo
- Output
- 1K image (PNG/JPEG download)
- Max resolution
- 1K only
- Credits
- 25 credits per image
See the edit in one drag
Muse Image is built for real photo edits — swap backgrounds, remove clutter, or restore old pictures with a single natural-language instruction.
Drag the slider to compare before and after
Example outputs
Change background
Try →“Replace the background with a soft sunset beach, keep the subject unchanged”
Multi-reference composite
Try →“Combine the person, product, and room from the reference photos into one natural scene”
Restore old photo
Try →“Restore faded colors, remove scratches, and sharpen facial details”
What you can do with Muse Image
Text to image — describe a scene and generate a visual from scratch in seconds.
Image edit — upload a photo and change backgrounds, remove objects, or restore old pictures with natural language.
Multi-reference — combine people, products, and environments from separate photos into one coherent image.
Instruction following — complex, multi-step prompts with stronger fidelity than single-pass generators.
Best for
Photo editing, multi-reference compositing, and natural-language changes to uploaded images. Start here when you have a real photo to edit.
Related
Is Muse Image the same as Meta AI?
No. Muse Image is the model name. Meta Image is our independent platform that runs it with credits, billing, and history. This is not Meta Platforms' official Meta AI app.
How many credits does Muse Image cost?
25 credits per image at 1K resolution. Muse Image supports 1K output only on Meta Image.
Third-party models
Model names refer to third-party AI integrations available through Meta Image. Meta Image is an independent platform — not affiliated with or endorsed by Meta Platforms, Inc. or any model provider.