Nvidia Unveils Lightning-Fast Open AI Model for AI Agents

Nvidia has released Nemotron 3.5 Lightning, an open AI model designed to be a fast workhorse inside long-running AI agent systems.

The new model activates 3 billion parameters per token and supports context windows of up to 1 million tokens.

Nvidia claims that Nemotron 3.5 Lightning can deliver up to four times the output speed of similar-sized models, thanks to its mixture-of-experts architecture and additional features such as multi-token prediction and speculative decoding.