Nemotron 3.5 Lightning is an ultra-fast open-weights model balancing massive efficiency with strong intelligence. The 3.6-billion-parameter model matches much larger systems on benchmark indices while outputting nearly 670 tokens per second. This release highlights Nvidia's strategic shift toward highly optimized, low-latency models designed for rapid inference applications.
Opening Kapyn…