Nvidia has expanded its AI portfolio with the launch of Nemotron 3.5 Lightning, a 30-billion parameter Mixture-of-Experts model, alongside NeMo Switchyard, an open-source library for intelligent model routing designed to optimize developer workflows and costs.
The new model excels at code reviews, tool usage, and security monitoring, offering output speeds up to four times higher than its peers. It processes agentic tasks 30 percent faster and runs locally on GeForce RTX PCs, DGX systems, and Jetson hardware.
NeMo Switchyard enables routing based on quality, latency, and cost, with LangChain reporting 74 percent lower costs on multi-step tasks. Both tools are now available via GitHub, Hugging Face, and Nvidia, providing scalable solutions for enterprise AI needs.