Meta Muse Image

Muse Image is Meta's image model that works like an agent. It plans a multi-part prompt before it draws, can search the web for facts, and edits or combines up to 10 reference images at about 2.5 megapixels.

Make pictures with Meta Muse Image Price per picture

What it is good at

Following long, multi-part prompts: it reasons before it renders and refines its own result.
Reference work: edits an image, combines several images, and keeps style and subject steady across a series.
Text inside images and factual subjects, where it can look things up with web search.

Honest limits

Only the aspect ratio is used: every picture is rendered at about 2.5 megapixels, so you cannot choose a larger size.
Edit mode takes 1 to 10 input images, and one request makes up to 10 pictures.
The ratio must be between 1:16 and 16:1.

Meta Muse Image price and free credits

From 3 credits a picture. The same quote the generator uses.

Credits for one picture (text to image; dollars at the Starter plan rate of $12 for 1,000 credits)
Size or quality	One picture
One picture	3 credits ($0.036)

Price on Ziztu. One picture costs 3 credits ($0.036) (one picture). Bigger plans and packs cost less per credit.
Free credits on Ziztu. A new account gets 40 free credits, no card, for scripts, character art, storyboards, voices and sound. Meta Muse Image is one of the image models open to the free credits, so you can make pictures with it before you pay. Video starts on a plan, from $12 a month for 1,000 credits.
Free from Meta. Meta may give free use of Meta Muse Image in its own apps. Those allowances, limits and watermark rules are set by Meta, change often and are not Ziztu's, so read Meta's own page before you count on one.

Prompts to try on Meta Muse Image

Copy a prompt, or press Try it to open the generator with the model and the prompt in place.

A poster for a neighborhood jazz night. A woman in a yellow dress sings into a vintage microphone. Large headline "JAZZ ON ELM STREET", smaller line "Friday 8 pm", warm red and gold palette.
Makes: A poster with the exact words in place.
Try it
A character sheet of a girl with short black hair, a green raincoat and white boots: front, side and back views on a plain light background, flat even light.
Makes: A reference sheet for later video clips.
Try it
A bright cafe interior with a wooden counter, hanging plants and a chalkboard menu, morning sun through large windows, no people.
Makes: A location plate for a scene.
Try it
A bicycle courier with a mint green helmet stopped at a crosswalk in a sunny city, shallow depth of field, photo style.
Makes: A first frame for image to video.
Try it
A cheerful infographic about how a bean becomes a cup of coffee, five steps with simple icons and the title "From bean to cup".
Makes: A diagram with text inside the picture.
Try it
An old fisherman with a white beard and a blue knit cap mending a net on a wooden dock at sunrise, warm light on his hands.
Makes: A portrait that can serve as a character reference.
Try it

Prompts for every model

Prompt tips for Meta Muse Image

Write the full brief in one prompt, with the subject, the setting, the style and any words that must appear.
To refine a picture, pass the last result back with a short instruction such as "make the sky warmer".
For a consistent series, give the same reference image each time and describe only what changes.

Meta Muse Image in Ziztu

Meta Muse Image makes the pictures that other models animate. In the AI Video Generator open Reference to video, then Make pictures, pick Meta Muse Image, and use what it makes as the first frame or the references of a clip. In Ziztu Studio the same pictures become character sheets that travel with every shot.

Other image models

Make the same picture on a second model and keep the one you like.

ByteDance Seedream 5.0 Flash Seedream 5.0 Flash is ByteDance Seed's fast, low-cost image model for making and editing pictures. Edit mode takes up to 10 reference images, and one request makes up to 6 pictures at 1K or 2K sizes. From 7 credits a picture0 OpenAI GPT Image 2.5 GPT Image 2.5 is OpenAI's image model in two tiers: Flare for fast everyday pictures and Sunburst for precise edits. People search for it as GPT Image 2 or GPT Image 2.5. Ziztu runs GPT Image 2.5. From 2 credits a picture1 Google Nano Banana Pro Nano Banana Pro is Google's image model built on Gemini 3. It makes and edits pictures at 1K, 2K or 4K, renders clear text, keeps the same people across several images, and can use web search for up-to-date facts. From 38 credits a picture2 xAI Grok Imagine Image 2.0 Grok Imagine Image 2.0 is xAI's image model for making pictures from text and editing up to 5 reference images. It offers 1k and 2k sizes, low and medium quality modes, and 13 aspect ratios. From 15 credits a picture3 Ideogram Ideogram 4.5 Ideogram 4.5 is the image model for posters, logos and accurate text in pictures. Edit mode changes part of an image and leaves the rest alone, with up to 4 reference images, an optional mask and a high-precision setting. From 8 credits a picture4

Frequently asked questions

What is Meta Muse Image?

How much does a Meta Muse Image picture cost?

On Ziztu one picture costs 3 credits ($0.036) (one picture). Dollars are at the Starter plan rate of $12 for 1,000 credits; bigger plans and packs cost less per credit.

Is Meta Muse Image free to use?

Free credits on Ziztu. A new account gets 40 free credits, no card, for scripts, character art, storyboards, voices and sound. Meta Muse Image is one of the image models open to the free credits, so you can make pictures with it before you pay. Video starts on a plan, from $12 a month for 1,000 credits. Free from Meta. Meta may give free use of Meta Muse Image in its own apps. Those allowances, limits and watermark rules are set by Meta, change often and are not Ziztu's, so read Meta's own page before you count on one.

How do I use Meta Muse Image for reference pictures?

Open the AI Video Generator, choose Reference to video, open Make pictures, pick Meta Muse Image as the image model, describe the person, place or prop and make one to four pictures. Pick the ones you like as the references of a clip.

Can Meta Muse Image keep the same character across pictures?

Reference work: edits an image, combines several images, and keeps style and subject steady across a series. For a full episode, Ziztu Studio keeps character sheets and sends them with every shot.

Is Meta Muse any good? A short review

What the maker's pages say it does well: it plans a long, multi-part prompt before it draws, can look up facts with web search, and edits or combines up to 10 reference pictures at about 2.5 megapixels. What it costs is on this page. How a picture looks is a matter of taste, so the quickest review is to make the same picture on Muse and on a second image model and keep the better one.

Is there a Meta Muse Image API?

Ziztu is the app, not an API: it does not sell API access to Meta Muse Image, and the generator is not a public endpoint. Developers can call Meta Muse Image through the platforms that host it. The endpoint IDs from the facts we read on 15 October 2026 (sources are listed at the bottom of this page) are meta/muse-image/text-to-image, meta/muse-image/edit. Check the host's own page for its price and terms.

Sources

Make pictures with Meta Muse Image and see the price first.

40 free credits, no card, for pictures, scripts and voices. Video starts on a plan from $12 a month.

Try it free