OllamaFrom the builder
NVIDIA Nemotron 3.5 Lightning
NVIDIA's open model Nemotron 3.5 Lightning is now on Ollama and runs fully on local devices, aimed at agents that keep running, call tools and work through multi-step tasks. It uses a hybrid Mixture-of-Experts design with 30B total and 3B active parameters per token, and supports up to 1M tokens of context. According to the post, it delivers 4x the throughput and 30% faster task completion than leading open models of similar size.
Most of these steps don’t need a large model. At 3B active parameters per token, it’s built for local systems rather than the datacenter, and running locally means your data stays on your device.
Saved . A shortened excerpt from the official blog; the original publisher retains authorship.