NVIDIA Nemotron-3.5 Lightning-30B-A3B-NVFP4 Model Card: Open Weights, 1M Context, and Multi-Backend Deployment
Summary
NVIDIA's Nemotron-3.5 Lightning-30B-A3B-NVFP4 is a 30B total parameter LLM with 3B active parameters, featuring a Mixture-of-Experts hybrid architecture and an extended 1M token context. The model card documents multiple deployment paths (vLLM, TensorRT-LLM, SGLang) and open-weight licensing under OpenMDW-1.1, with performance benchmarks and a detailed quick-start guide. Released Aug 11, 2026, the card positions Nemotron-3.5 as a commercial-ready AI agent/automation tool for developers building AI agents, chatbots, and RAG workflows.