Guides / Images
How to Generate Images with Muse: A Step-by-Step Tutorial
Muse’s image generation is powered by Meta’s Muse Image model, which ranked second on Arena’s text-to-image leaderboard when it launched in July 2026. It’s good at following instructions and at writing readable text inside images. The quality you get depends mostly on how you describe what you want, so this tutorial focuses on writing prompts that work.
Step 1: Ask for an image
In a Muse chat, describe the image you want. You don’t need special commands. Starting with “Create an image of…” or “Make a picture of…” makes your intent clear. Muse will generate the image in the conversation, and you can save it from there.
Step 2: Use the five-part prompt formula
Good image prompts cover five things. You don’t need all five every time, but the more you include, the less Muse has to guess.
- Subject: who or what is in the image, and what they’re doing.
- Setting: where it takes place, and when.
- Lighting: the time of day and quality of light, such as soft morning light or harsh midday sun.
- Style: photo, illustration, 3D animation, watercolour, and so on, plus camera angle or lens if it’s a photo.
- Constraints: aspect ratio and anything to leave out, like “no text” or “no people in the background”.
Worked example
Here’s a weak prompt and a stronger version of the same idea.
The first is a list of keywords, which tends to produce generic results. The second reads like directions you’d give a photographer: it says exactly who, where, when, how it’s lit, and how it’s framed.
Write like you’re talking to a photographer
Muse understands full sentences well. A useful test is to read your prompt out loud. If it sounds like a natural instruction, you’re on the right track. If it sounds like a list of tags, rewrite it.
Keep each element to two or three describing words. Piling on adjectives like “beautiful, stunning, epic, breathtaking” doesn’t make the image better. It tends to blur the result into something bland.
Choosing an aspect ratio
Muse often defaults to a tall, phone-friendly format. If you need something else, say so in the prompt:
- 16:9 for YouTube thumbnails, video frames and presentations
- 9:16 for Stories, Reels and Shorts
- 1:1 for profile pictures and grid posts
- 4:5 for portrait feed posts
If the result still comes out in the wrong shape, ask Muse to redo it at the ratio you need, or crop it afterwards.
Putting text in images
Muse Image is noticeably better than many older image tools at spelling words correctly, which makes it handy for posters, menus, invitations and thumbnails. Put the exact words in quotation marks and say where they go:
Keep text short. A few words come out reliably, while full paragraphs are more likely to contain mistakes. Always check the spelling before you publish.
Using reference images
You can upload photos for Muse to work from, such as a product, a person’s outfit, or a background. If you upload several, Muse can combine them into one scene. When one reference gets ignored, upload the images one at a time and label each one, for example “This is the product” and “This is the background”. Our editing guide goes into more detail.
Common mistakes
- Keyword stuffing. Long tag lists produce generic images. Write sentences.
- Forgetting the aspect ratio. You’ll get a phone-shaped image when you wanted a thumbnail.
- Regenerating instead of editing. If an image is 80% right, ask Muse to fix the 20% instead of starting over. It’s faster and uses fewer tokens.
- Not checking details. Look closely at hands, text and background objects before you use an image.
- Using an old app version. Update the app so you’re on the newest model.
Practice prompts
Try these to get a feel for how Muse responds, then adapt them. You’ll find many more in our prompt templates.