Flux-2-Flex (Official format)

FLUX.2 Flex is a flexible and efficient image generation and editing solution series launched by WaveSpeedAI. It consists of two collaborative models—Text-to-Image and Edit—and is built on a unified architecture that is lightweight, open-source, and LoRA-friendly. This architecture enables a complete end-to-end image creation workflow—from creative generation and style exploration to final output and fine-tuning optimization.

• Two core capabilities: It features versatile text-to-image generation capabilities as well as high-precision editing of existing images;
• Quality-speed tradeoff: In the generation phase, users can adjust the number of inference steps and the guidance scale, covering everything from rough sketches to near-final outputs. In the editing phase, users can control the number of inference steps as needed, making it suitable for scenarios ranging from simple retouching to complex edits;
• Precise and controllable operations: The editing phase supports natural language instructions and allows precise color control via hexadecimal color codes, without disrupting the original image’s core composition or main subject;
• Wide-ranging style adaptation: It can generate images in diverse styles, including product photos and illustrations, and also enables batch-style consistency across multiple images;
• Flexible and extensible architecture: The open architecture supports deep customization, is LoRA-friendly and allows custom training, enabling users to lock in unique styles or themes;
• Optimized resource costs: It strikes a balance between resource consumption and performance, making it ideal for batch processing and cost-sensitive creative workflows;
• Stable and controllable results: It supports seed control to ensure generation consistency and allows adjustable editing guidance intensity, meeting the needs of iterative design and precise output;

Price：Pricing | Black Forest Labs