Black Forest Labs has released FLUX.1 Kontext [dev], an open weights image editing model designed to deliver proprietary-level performance on consumer hardware. The 12-billion parameter model offers best-in-class image editing capabilities and is available with self-serve commercial licensing. It has been integrated into popular platforms such as Hugging Face Inference Providers and Fireworks, enabling users to perform advanced image editing tasks including character consistency, local editing, style referencing, and panoramic image creation at 1024-pixel resolution with consistent lighting and minimal object cut-off. Additionally, the model runs efficiently on hardware like Groq's Kimi K2 accelerator, which has demonstrated impressive speed and adherence to prompt instructions, facilitating applications such as multi-agent research clones and rapid coding workflows. The FLUX.1 Kontext model and its LoRA variants are fully open source, supporting infinite canvas editing and broad accessibility for developers and researchers. This release represents a notable advancement in open-source generative image and video models, offering competitive alternatives to proprietary solutions while reducing compute and data requirements.