What is multi-reference editing?

Combine elements from multiple source images (up to 8-10 depending on model) into one cohesive output.

Use cases:

- Fashion editorials with consistent characters

- Product mockups in various contexts

- Interior design (combine furniture from multiple photos)

- Character consistency across variations

Limits:

- [pro] API: Up to 8 images (9MP total limit)

- [pro] playground: Up to 10 images

- [flex]: Up to 10 images

- [dev]: Recommended max 6 images

How do I reference specific images when editing with FLUX.2?

Natural language:

"The man is leaning against the wall. The woman is walking past him, carrying the tote bag and wearing the black boots."

FLUX.2 identifies elements across images based on your descriptions.

Explicit indexing:

"The person from image 1 is petting the cat from image 2, the bird from image 3 is next to them"