Skip to main content
The engine you select determines which FLUX.2 model runs your generation, how the prompt is formatted, how many images you get, and how many credits it costs.
FLUX.2 Pro — 10 credits, 8 images per generationPro sends a JSON-structured prompt to FLUX.2 Pro. The structured format lets the model anchor specific objects (@product1, @model, etc.) more precisely than prose can.
Example Pro engine result showing a detailed product photo with clean lighting
Use Pro when:
  • You’re producing final or hero images
  • Product detail and material accuracy matter
  • You want the maximum number of variations in one run (8 images)
Pro costs more per generation but produces more images, so the per-image cost is comparable to Fast.

​
Engine comparison

EngineModelCreditsImagesPrompt format
ProFLUX.2 Pro108JSON structured
FastFLUX.2 Klein34Flowing prose
Studio UploadFLUX.2 Pro141JSON + Learn metadata

​
Choosing an engine

A common workflow: use Fast to explore and find the right combination of references and settings, then switch to Pro for a high-quality final batch. Use Studio Upload when you need to match an established product identity precisely.