Nemotron-3.5-lightning-30B-a3B Model by Nvidia for use on 1 GPU

Nvidia released the Nemotron-3.5-lightning-30B-a3B model, a 30-billion-parameter AI model with 3 billion active parameters, designed to run on a single GPU. The model is available via Nvidia's NIM API, supporting a context length of 128,000 tokens and a maximum output of 16,384 tokens, with features like reasoning and thinking enabled. It is part of Nvidia's push to provide efficient, high-performance models for enterprise AI applications.

Skip to main content Explore Models Skills Blueprints GPUs Docs Search ⌘K Ctrl+K ? Help Center Getting Started 1 Set up your account Create and verify your account to unlock full access to NVIDIA NIM APIs. Create an Account 2 Generate API Key 3 Make your first API call 4 Prototype in your environment 5 Connect to inference partners Resources Developer Forums Contact Support FAQs Login nemotron-3.5-lightning-30b-a3b Model by NVIDIA | NVIDIA NIM nvidia / nemotron-3.5-lightning-30b-a3b Build Build Playground Playground Model Card Model Card API Reference Prototype Start building with a free API endpoint. Python LangChain Node Shell Generate API Key Copied python from openai import OpenAI client = OpenAI base url = "https://integrate.api.nvidia.com/v1", api key = "$NVIDIA API KEY" completion = client.chat.completions.create model="nvidia/nemotron-3.5-lightning-30b-a3b", messages= {"role":"user","content":""} , temperature=1, top p=0.95, max tokens=16384, extra body={"chat template kwargs":{"enable thinking":True},"reasoning budget":16384}, stream=True for chunk in completion: if not chunk.choices: continue reasoning = getattr chunk.choices 0 .delta, "reasoning content", None if reasoning: print reasoning, end="" if chunk.choices 0 .delta.content is not None: print chunk.choices 0 .delta.content, end="" Deploy Ready to scale? Choose your deployment path. Partner Endpoints Self-Hosted Deployments Available Integrations Deploy this model now on your endpoint provider of choice