A massive mixture-of-experts model that activates only 17 billion of its 397 billion parameters per forward pass, giving it the computational footprint of a mid-sized model while drawing on a vast pool of specialized knowledge. It handles complex reasoning, coding, and multilingual tasks with notable depth, and the NVFP4 quantization means it runs efficiently on NVIDIA hardware without dramatic quality loss.