Control image generation

Guide image generation with more than just text. Use edge detection, depth maps, and sketches to get the results you want.

What you can do

Use edge detection to keep specific shapes and lines from an image. Great for preserving architecture or recreating specific compositions.

Control depth and space with depth maps. Perfect for retexturing objects while keeping their 3D structure intact.

Turn rough sketches into detailed images. Start with a simple drawing and let the model fill in the details.

Create artistic QR codes and patterns that actually work when scanned.

Models we recommend

For professional work

The FLUX models deliver reliable results with clean outputs:

• FLUX.1 Depth Pro for depth-based control
• FLUX.1 Canny Pro for edge detection

Learn more about FLUX Tools →

For speed

Latent Consistency Model runs in about 0.6 seconds per image. It includes:

• Edge detection control
• Image-to-image generation
• Support for large batches (up to 50 images)
• Quality results in just 4-8 steps

For creative projects

Illusion makes cool images with spiral or other patterns, and artistic QR codes that still scan.

ControlNet Scribble turns simple drawings into detailed images - great for sketching out ideas.

Try it out

Play with different control methods in the playground. Test models side by side and experiment until you find what works.

Open the playground →

Want to learn how it works? Read our guide →

Questions? Join us on Discord.