Flux AIFlux AI Documentation
Flux AI docs for model switching, prompt writing, and AI image generation workflows.
What this documentation covers
These docs are intentionally narrow. They focus on the parts users actually need when using an AI image generation site:
- how to switch models
- when to use text-to-image or image-to-image
- how to write stronger prompts and negative prompts
- what each model tends to generate well in real workflows
Read this first
Model Switching Guide
Learn when to use Flux Schnell, Flux Dev, Flux 2 Pro, Flux 2 Max, Qwen Image, Seedream 4, or Nano Banana Pro.
Prompt Optimization Guide
Learn how to write subject, material, lighting, composition, and negative prompts in a cleaner order.
Model Pages
Open long-tail landing pages for each model and match the right model to the right search intent.
FAQ
Get direct answers about prompt length, image-to-image, quality, and model fit.
Practical workflow
Use the site in this order:
- Start with
Text-to-Imagewhen you only have a prompt. - Switch to
Image-to-Imagewhen you already have a base image and want to preserve layout or structure. - Pick a model based on output goal: speed, prompt control, detail, product photography, or premium finish.
- Rewrite the prompt until the subject, materials, lighting, composition, and exclusions are explicit.
What the models are usually good at
Nano Banana Pro: minimalist product photography, tabletop still life, glassware, bottles, and quiet commercial composition.Qwen Image: longer prompts, bilingual prompts, scene structure, and stronger prompt-following.Seedream 4: sharper detail, cleaner textures, polished surfaces, and high-resolution commercial visuals.Flux Schnell: quick drafts, fast A/B prompt tests, and early concept exploration.Flux Dev: balanced day-to-day generation with stronger control for iterative work.Flux 2 Pro/Flux 2 Max: premium output, stronger subject consistency, and more polished final campaign visuals.