Black Forest Labs has released FLUX.1 Kontext, a new family of AI models that can both generate and edit images. The German startup announced the launch on Thursday, introducing technology that accepts both text prompts and reference images as input.
The company offers two main versions of the model. FLUX.1 Kontext [pro] allows users to generate images and refine them through multiple editing rounds while preserving characters and styles. FLUX.1 Kontext [max] focuses on speed and prompt accuracy. Black Forest Labs claims the models work up to eight times faster than current leading alternatives.
Unlike traditional text-to-image generators, Kontext models can perform what the company calls “in-context generation.” This means they can modify existing images based on text instructions, such as changing clothes, restyling scenes, or rewriting text within images.
The models use flow matching technology rather than diffusion, which Black Forest Labs says provides more flexibility. Flow models learn from continuous data streams and define paths between noisy and useful information, differing from the denoising approach used by many competitors.
Users can test the models through Black Forest Labs’ new playground platform, which provides 200 free credits for new users. The models are also available through third-party platforms including Replicate, KreaAI, and LeonardoAI.
Black Forest Labs was founded by former Stability AI team members who created Stable Diffusion. The company reportedly sought $100 million in funding at a $1 billion valuation last year.
Sources: TechCrunch, VentureBeat