NVIDIA has partnered with Black Forest Labs to optimize the 'FLUX.1 Kontext' model for RTX GPUs to maximize the efficiency and quality of AI image generation.

This collaboration is attracting attention as a key technology that streamlines the image creation workflow while enabling high-quality visuals and a real-time editing environment.

Flux.1 Context is a context-based open-weight model that supports both text and image prompts, enabling image creation and editing using only intuitive natural language input.

It is characterized by the ability to perform flexible style changes and precise detail adjustments based on a reference image without complex masks, depth maps, or edge maps.

NVIDIA has significantly reduced VRAM requirements by optimizing models through TensorRT and quantization technology.

This enables stable fast inference and iterative editing on RTX GPU-based devices and is immediately available for use in ComfyUI and Playground.

In addition, an NVIDIA NIM microservices version is also scheduled to be released in August.

The model's flexibility and control stand out with the following features.

Character consistency is It maintains unique characteristics across various scenes and angles, and custom editing allows you to selectively modify only certain elements of the image. Style switching can apply the mood of a reference image to a new scene, and through real-time execution, it supports rapid feedback and iterative generation with minimal delay.

Model weights and TensorRT variants are available for free download via Hugging Face, and accessibility has been ensured so that they can be run in a local environment by reducing Ada-based FP8 to 12GB and Blackwell-based FP4 to 7GB.

In particular, SVDQuant technology, optimized for the NVIDIA GeForce RTX 40/50 series, provides efficient processing speeds while maintaining high image quality.

In addition to AI image generation, Google's recently unveiled small language model 'Gemma 3n' also delivers optimal performance on NVIDIA RTX and Jetson platforms.

Gemma 3n works with Ollama and Llama.cpp-based applications to provide RTX-accelerated natural language processing capabilities, offering a wide range of application opportunities for AI enthusiasts and developers.

In addition, NVIDIA is hosting the 'Project G-Assist Plugin Hackathon' online until July 17, providing developers with the opportunity to create custom AI plugins.

A G-Assist webinar is also scheduled for July 10, where you can exchange various information through real-time Q&A.

NVIDIA is accelerating community-driven technology sharing and content creation support through its Discord server and RTX AI Garage blog series, and is also continuously providing information for next-generation creative innovation through AI PCs and workstations.

NVIDIA's latest AI technologies, including Flux.1 Context, can be found on NVIDIA's official blog, social media channels, and HuggingFace, and You can also receive the latest news through related newsletters.