Features · Make
A sentence, a shape, a count.
Say what, say how big, press Generate.
The composer.
Write a prompt. Choose a shape: square, classic, portrait, photo, poster, wide, tall or cinema. Choose one to four pictures. Add a seed to repeat a result, and a negative prompt where the model accepts one. The composer names the model that will make the run, and a run appears in the history at once.
- Making pictures needs Premium or Max
- A run is counted when its pictures are made
- A failed picture is not counted
Eight shapes.
Square for posts and avatars, classic and portrait for slides and prints, photo and poster, wide and tall for screens and stories, and cinema for banners.
- 1:1, 4:3, 3:4, 3:2, 2:3, 16:9, 9:16, 21:9
- A model may support a subset
- The longest side is limited by the model
Questions
About making pictures.
Which shapes are there?
Eight: 1:1, 4:3, 3:4, 3:2, 2:3, 16:9, 9:16 and 21:9. A model may support fewer.
How many pictures at once?
One to four a run.
What is a seed?
A number that repeats a result. With the Preview, the same words and seed give the same picture.
Why does a run wait for an open page?
The Preview runs in a browser. A run queued by someone else, such as your AI, is made as soon as somebody in the workspace has AI Pics open.
Is a failed picture counted?
No. A run is counted from the pictures it made.
How many pictures can I make?
Premium up to 2,000 a month and Max up to 5,000. Pro makes none.
Which files do I get?
The pictures are files you can download any time. The format depends on the provider that made them.