01
Renders legible, accurate text inside images. Strong on logos, poster headlines, and multi-word body copy
02
Handles complex, lengthy text compositions with improved typesetting
03
Supports 15+ named aspect ratios: 1:1, 9:16, 16:9, 2:3, 3:2, 4:5, 5:4, 4:3, 3:4, 1:2, 2:1, 1:3, 3:1, 10:16, 16:10
04
Explicit resolution overrides from 512x1536 up to 1536x640, with many intermediate sizes. Resolution overrides aspect ratio when both are set
05
Upload up to 3 style reference images to guide generation aesthetics
06
60+ named style presets for V3 models: 80s Illustration, Art Deco, Bauhaus, Cubism, Double Exposure, Oil Painting, Watercolor, Woodblock Print, and more
07
Style Codes let you save and reuse any generated style across future prompts
08
Random style feature draws from 4.3 billion style presets
09
Inpainting support: supply a source image plus a black-and-white mask. Black pixels are replaced. White pixels are preserved
10
Magic Prompt auto-optimizes your input to maximize variety and output quality
11
Accepts prompts written in multiple languages via the Magic Prompt toggle
12
Choose from: None, Auto, General, Realistic, or Design
13
Synchronous API. Image links return immediately and must be downloaded promptly. URLs are time-limited
14
Outputs are cleared for commercial use including logos, ads, and client work
15
Inputs and outputs are not retained and not used for model training