Readable text rendering
Posters, covers and banners with legible typography — quote the exact words and specify weight, case and alignment.
Turn a prompt into a polished image, or follow plain-language instructions to edit and combine up to 10 reference images — up to 4K, nine aspect ratios and built-in prompt enhancement.
Black Forest Labs’ lightweight FLUX 3 image model — fast text-to-image and multi-reference editing with readable text rendering and precise color control.
Low per-image cost and multi-reference editing for production-ready visuals.
Posters, covers and banners with legible typography — quote the exact words and specify weight, case and alignment.
Combine products, characters and scenes from up to 10 reference images into one coherent picture.
Swap clothing, light or place while the face, hair and details stay put — describe what to change and what to keep.
Include six-digit hex codes directly in the prompt for branding, backgrounds and garments.
1K drafts finish in seconds at low cost — perfect for generating many variations before choosing a direction.
Short prompts are rewritten into fuller descriptions by default; turn it off for literal instructions.
Pick a mode, configure the output and generate.
Text to Image for new visuals, or Edit Image with 1–10 references for edits and composites.
Describe the subject, lighting and materials — or the change you want, naming what must stay unchanged.
Set aspect ratio, resolution (1K–4K) and format (PNG, JPEG or WebP).
The studio estimates your cost live. Preview the result inline and download the finished image.
Answers to common questions about inputs, settings and pricing.
Flux 3 Klein is Black Forest Labs’ lightweight FLUX 3 image model. It turns a prompt into a polished image and follows plain-language instructions to edit and combine up to 10 reference images in a single request.
Text to Image generates a new image from a prompt with your chosen aspect ratio. Edit Image takes 1 to 10 reference images and follows an instruction — swap clothing, restyle a scene, or composite subjects from several images into one picture.
Refer to them by order in natural language, for example “put the hat from image 2 on the person in image 1”. Reference images keep their upload order.
Choose 1K, 2K or 4K output (roughly 1, 4 and 16 megapixels). Nine aspect ratios are supported from 21:9 ultra-wide to 9:21 phone strips — Edit mode also offers auto, which matches the first reference image.
Cost is a per-image rate by resolution: 8 gems per 1K image, 16 per 2K image and 32 per 4K image. Reference images in Edit mode are free.
On by default — it rewrites short prompts into fuller descriptions, which noticeably improves quality. Turn it off when your instruction must be followed literally, especially for precise edits.
Reference images: JPG, PNG or WebP up to 30MB each. Output: PNG, JPEG or WebP.
Start with a prompt or a set of references and generate polished images with Flux 3 Klein.
Start creating with Flux 3