Every AI tool answers "I need more variations" with "generate more." More credits, more GPU time, more randomness, more waiting. The combinator answers differently: the variations already exist inside your batch. You just need to extract and reassemble them.
Upload your AI-generated character images. Draw freeform shapes on the swappable regions — head, outfit, accessories. The combinatorial engine computes every possible assignment: face from image 1 + outfit from image 4 + accessories from image 9. Then face from image 2 + outfit from image 4 + accessories from image 7. Every permutation. 9 images × 3 shapes = 620+ unique character variations.
When you search "combine AI images," every result shows you AI blending tools — neural networks that fuse two images into a new hybrid. That's a fundamentally different operation. Here's why it matters:
Feeds your images into a neural network. The AI interprets both and generates a NEW hybrid. Output pixels are hallucinated — they don't exist in either source. Result is unpredictable. Costs GPU time and credits. Good for creative exploration. Bad for precise character work where you need exact control over what goes where.
Extracts exact pixel regions from your existing images using freeform shapes. Reassembles them in every mathematical combination. No neural network. The face from image 3 is the ACTUAL face from image 3 — not an AI approximation. Deterministic: run it twice, get the same result. No GPU, no credits, no hallucination.
The distinction matters because AI blending introduces drift — the very problem you're trying to solve. If your character's face drifted across 10 generations, blending two of them produces a face that looks like neither. Combinatorial reassembly gives you the EXACT face from the image where it looked best.
Every AI image platform monetizes the generation loop. You need more variations → you generate more → you spend more credits. A combinator that produces 620+ variations from 9 existing images without generating is revenue they lose. They have no incentive to build it. Compix has no generation revenue to protect. That's why this exists here and nowhere else.
Step-by-step walkthrough of freeform region extraction, the compound math, and the complete workflow from source images to finished variations. See the guide →
The practical guide: take the face from image 1, the outfit from image 3, the boots from image 7. Every combination. How to combine →
Before combining, blink your batch to spot which regions drifted. The diff heatmap shows what's worth extracting. Detect drift →