Can Copilot or Meta AI Make Images? How to Use Both

Copilot can make images, and so can Meta AI, both inside apps you may already have. Microsoft's Copilot creates and edits pictures in its free version within usage limits, and Microsoft runs a separate free site, Bing Image Creator. Meta AI creates and changes pictures inside the Meta AI app and in WhatsApp chats. The steps for each follow, along with what the free versions allow and the rules that decide whether you can use the results for a business.

Can Copilot make images?

Yes. Copilot has a limited free tier, subject to available capacity, and Microsoft's plan page lists AI image generation on every paid tier at higher allowances than free. So the short answer to "does Copilot make images" is yes, on the web at copilot.microsoft.com and in the Copilot app. As of September 2026, the US plans and their image allowances are:

On the paid plans, the AI features belong to the person who owns the subscription, not to family members on it. Some features count uses in a set period and others draw on monthly AI credits as you use Copilot in apps such as Designer, Photos and Paint. You can check what's left under Services and subscriptions in your Microsoft account.

How to use the Copilot image generator

Microsoft has three routes, and each works a little differently.

In Copilot

1. Go to copilot.microsoft.com or open the Copilot app.
2. Describe the image in as much detail as possible, for example "Create an image of a paper lantern festival over a river at night, watercolor style, wide format."
3. Wait for the result. Microsoft says an image takes a couple of minutes.
4. Keep refining in that conversation, one request at a time.
5. To change a photo you took, select the Open icon (+), choose "Add images or files", pick the photo, and describe the change, such as "remove the cars from the street behind us".

Microsoft's advice is that more detail helps Copilot match a style, such as realism, comic book or abstract, along with the mood and setting you describe. Upload only photos that are yours or that you're licensed to use, and never ones that show people without their consent. Copilot keeps generated and uploaded images for no longer than 18 months, and you can remove them sooner by deleting them from your conversation history.

In the Microsoft 365 Copilot app

Microsoft 365 subscribers get a separate image form with set styles and shapes.

1. Open Create in the left pane, then choose "Create an image".
2. Write your prompt under "Describe your image", or pick a template to pre-fill one. Use "Add an image" to attach a reference picture.
3. Under Style, pick a look such as Photorealistic, Surrealism or Anime.
4. Choose a shape: Square, Portrait or Wide.
5. Select Create. Your images are saved in Copilot Library.

In Bing Image Creator

Bing Image Creator is Microsoft's free image site, separate from Copilot.

1. Go to Bing Image Creator and sign in with a personal Microsoft account. Work and school accounts can't use it.
2. Choose a model: MAI-Image-2.5-Flash or MAI-Image-2e. Editing an uploaded photo always uses MAI-Image-2.5-Flash.
3. Type a detailed prompt and create it. One prompt can return several images, which counts as one creation.
4. Count your fast creations. Microsoft's pages disagree on the daily number (the FAQ says 15, the sign-in panel says 10). After that you can spend Microsoft Rewards points on more fast ones, or keep going at the slower standard speed.
5. Plan around the FAQ's cap of 200 prompts in 24 hours. Each "Edit image" click counts toward it.

You can switch on a visible watermark that marks your images as AI-made, for the current visit, and write prompts in more than 100 languages. If your prompt closely matches one used recently, Image Creator may show a temporarily saved earlier result instead of drawing a new one.

Can Meta AI make images?

Yes. Meta AI makes pictures and short videos in the Meta AI app, and it creates and edits images inside WhatsApp. Meta says its apps and websites aren't available everywhere yet, and some features may be missing in your area.

How to use the Meta AI image generator in the Meta AI app

1. Open the Meta AI app.
2. Tap "Ask Meta AI…" and type your prompt. Meta recommends "imagine" or "create an image" as the first words, plus plenty of detail about the scene.
3. Tap Send. Meta AI replies with several pictures per prompt.
4. To put yourself in the picture, use "me" or "myself" in the prompt. To include someone else, type @ and their username.
5. To turn a result into a short video, tap the image, then Animate, and add a prompt or send it without one.

Meta AI can also edit a photo or video you upload. Meta labels AI-generated images made with its products. Its help center says it uses your chats with its AI to improve that AI, and that profile details such as age and interests personalize your answers.

How to use the Meta AI image generator in WhatsApp

WhatsApp builds Meta AI into chats, so you can make an image without leaving the app. In your chat with Meta AI, type a description of the picture and tap Send. In any other chat, including a group:

1. Open the chat, tap the paperclip, then tap AI images.
2. Tap New image, describe what you want, and tap the create icon. To change one of your own pictures instead, tap Use photo and pick it. You can also press and hold the microphone and describe the image out loud.
3. Tap refresh to see a new set, or use Edit to add, remove or change something, Restyle to try a different style, or Animate to make it move.
4. Tap Send to share it in the chat.
5. To keep a copy, press and hold it, then choose Save photo before sending or Save once it's in the chat.

WhatsApp suggests using the results as a group icon or a video-call background. Meta AI is only available in some countries, and it works in English, Arabic, French, German, Hindi, Indonesian, Italian, Portuguese, Spanish, Tagalog, Thai and Vietnamese. WhatsApp says your personal messages and calls stay end-to-end encrypted, and it offers an incognito chat with Meta AI that disappears once you leave it.

Can Copilot make images like ChatGPT?

Copilot, Meta AI and ChatGPT all make pictures from a written description, but they sit in different apps and come with different rules. As of September 2026:

ChatGPT has extras of its own, such as Sketch, which turns a rough drawing on your phone into a finished image; the walkthrough on how to make images with ChatGPT covers it. Claude, by contrast, doesn't draw pictures at all, as Claude's image abilities and limits explains.

Where Copilot and Meta AI images fall short

• Text in the picture. Microsoft's Designer FAQ cautions that words in its AI images can come out in the wrong language or misspelled. Check each word before it goes on a post or a print.
• Accuracy. WhatsApp's help center warns that some Meta AI images may be inaccurate or inappropriate, so check faces, hands and details before you share.
• Business use. Microsoft licenses Designer only for personal, non-commercial use. The Image Creator terms describe the service as for entertainment and say images may not be unique, since other people can get the same or similar results.
• Provenance marks. Designer attaches content credentials that show an image was made or edited with AI, and the Image Creator terms forbid removing them.
• Your uploads. Bing Image Creator keeps uploads for up to 30 days, deletes those with faces after that, and may keep the rest for up to 18 months.
• Availability. Meta and WhatsApp both say features vary by country and account, so a friend's WhatsApp may show options yours doesn't.

For the bigger question of who owns an AI picture and when you can sell it, see can AI-generated images be used commercially.

When a dedicated image tool fits better

A chat assistant is handy for a quick picture. When you want a set shape, reference photos and one place to keep your results, Image Generator in Toybox AI is a single form for that job. Write the description in "Describe your image", attach as many as 3 photos as a visual guide, and choose one of five shapes under "Size", including Story (9:16) for phone screens and Wide (16:9) for banners. For designs with words, flip the "Enhanced quality" switch inside "More options", then proofread each word.

Results you don't save are deleted after 24 hours, so save the ones worth keeping to your gallery. Toybox's privacy policy states that prompts, uploads and results aren't used for AI model training. The price is currently 50 credits a picture. Sign-up comes with 20 free credits, which isn't enough for one, and a 500-credit top-up currently costs $4.99 on the pricing page. To see how other generators stack up, read ChatGPT image alternatives side by side.