FLUX.1 Kontext

FLUX.1 Kontext is a vision-language model designed to generate images that are tightly grounded in detailed textual and visual context. It accepts long, structured prompts along with reference images, enabling it to produce outputs that accurately reflect complex scenes, styles, and constraints. The model is optimized for context-aware image synthesis, making it suitable for workflows where consistency, control, and adherence to specifications are critical.

Key capabilities include multi-image conditioning, where users can provide several reference images to guide composition, lighting, and aesthetics, and fine-grained text conditioning that respects attributes such as layout, objects, relationships, and style directives. FLUX.1 Kontext supports iterative refinement, allowing users to adjust prompts and references to converge on precise visual outcomes.