Flux.2 is a powerful text-to-image AI model that combines generation and editing in one unified system. Features fast generation, high-quality output, and specialized tools for typography and photo-realistic results.
Unified Generation & Editing
Create images from text or edit existing ones using the same unified model. The platform handles text-to-image generation and advanced editing (inpainting, outpainting) without switching tools, keeping lighting and style consistent across all your creations.
Up to 4 Megapixel Output
Generate high-resolution visuals up to 4 megapixels (approximately 2048x2048 or equivalent). The max quality variant delivers exceptional detail suitable for print, commercial use, and professional projects.
Fast Generation Mode
Our fastest model (klein variant) creates visuals in under a second. Perfect for rapid text-to-image generation, testing creative concepts, and high-volume workflows.
Multi-Reference Editing
Upload multiple reference images (like a face and a clothing style) and the engine blends them into new compositions. Supports up to 10 simultaneous references for consistent character design and style control across your creations.
Text Rendering
Create visuals with readable, accurate text. The flex variant is optimized for typography, generating complex posters, book covers, and logos with perfect spelling and font consistency. No manual text fixes needed.
RAW Realism Mode
Generate photorealistic visuals that mimic real camera sensor output, including natural noise, color grading, and optical imperfections. This mode reduces the typical artificial AI look for authentic, photographic results.