There are two places to add a file: the studio, which uses your photo, video or audio to make a result, and image to prompt, which reads a picture to write a prompt.
Studio uploads
You need to be signed in to upload to the studio. Photos can be PNG, JPEG or WebP. Studio tools that work with video or audio, such as lip sync, accept those files too, and the upload box says what it takes.
| File | Largest upload |
|---|---|
| Photos and images | 12 MB |
| Video | 60 MB |
| Audio | 25 MB |
Uploads are stored privately in your account and sent only to the model that runs your job.
Image to prompt
Image to prompt takes a JPG, PNG or WebP image under 4 MB and works without an account. The image is sent to the text model that writes the prompt and is not saved to your account. Each image counts as one run of your free daily prompts. See Using the prompt tools.
Getting a good result
- Use a sharp photo where the face or subject fills most of the frame.
- Face the camera with even light and nothing covering the face.
- For tools that take two people, use one clear photo of each person.
Input photo quality has more on light, framing and the photos each kind of tool needs.
Whose photos you may upload
Upload a photo of a person only if it is you or you have their permission. Photos of anyone under 18 may be uploaded only by a parent or legal guardian. The full rules are in What you can create.
Who can see your uploads
Your uploads are never shown to other people. Results appear in the public gallery only if you press Share on them, and you can make them private again. The privacy policy lists which providers process your files, and Data retention explains how long they are kept and how to have them deleted.