Powered by state-of-the-art AI text-to-image models

Text to Image

Optional. To make AI art of yourself, your kids, family or friends, first press New character below. You'll train your own LoRA.

Add your own characters (LoRAs)

Advanced

AI model

Default. ⭐️:3/3 💰: 2/3

⭐️:3/3 💰: 1/3

🏆 ⭐️:3/3 💰: 3/3

⭐️:2/3 💰: 1/3

⭐️:3/3 💰: 2/3

⭐️:3/3 💰: 3/3

⭐️:2/3 💰: 2/3

Aspect ratio

Quality

Resolution

Reasoning effort

Generate image in response

Size (aspect ratio)

Lower values mean more compression and lower quality.

Output JPG compression level

Output format

How much to adhere to the prompt? A higher value adheres to prompt more

Guidance scale

Height (pixels)

Hide Watermark

Seed

Negative prompt (what to keep out)

Number of images

Number of inference steps

Generate privately

Describe your image (prompt)

For a more details, but more imperfections, try 3 steps. For more refinement, try 10.

Refiner steps

This is the AI middleman that may improve your prompt and actually interacts with the GPT Image generator.

AI orchestrator model

i30 💰:1.0

i66 💰:75.0

i70 💰:125.0

🏆 i67 💰:25.0

i66 💰:12.5

i57 💰:37.5

i49 💰:15.0

i50 💰:1.3

i61 💰:75.0

i55 💰:37.5

i48 💰:11.3

🏆 i44 💰:25.0

i40 💰:5.0

i39 💰:100.0

Ground with Google Search

Transparent background

Auto-trim edges

Run your prompt through Magic Prompt before generating

Creates a second image with fewer imperfections.

Use SD XL Refiner. Creates a second image.

Width (pixels)

Results

AI art of Prompt: Infrastruce project by stabilityai/stable-diffusion-3-medium

AI Images of Prompt: Infrastruce Project

Results

Request another