Skip to main content

Search⌘KCtrl+K<br>?Help Center

Getting Started

1Set up your accountCreate and verify your account to unlock full access to NVIDIA NIM APIs.Create an Account

2Generate API Key

3Make your first API call

4Prototype in your environment

5Connect to inference partners

ResourcesDeveloper ForumsContact Support<br>FAQsLogin

nemotron-3.5-lightning-30b-a3b Model by NVIDIA | NVIDIA NIM

API Reference

PythonLangChainNodeShell

Generate API KeyCopiedCopy

from openai import OpenAI

client = OpenAI(<br>base_url = "https://integrate.api.nvidia.com/v1",<br>api_key = "$NVIDIA_API_KEY"

completion = client.chat.completions.create(<br>model="nvidia/nemotron-3.5-lightning-30b-a3b",<br>messages=[{"role":"user","content":""}],<br>temperature=1,<br>top_p=0.95,<br>max_tokens=16384,<br>extra_body={"chat_template_kwargs":{"enable_thinking":True},"reasoning_budget":16384},<br>stream=True

for chunk in completion:<br>if not chunk.choices:<br>continue<br>reasoning = getattr(chunk.choices[0].delta, "reasoning_content", None)<br>if reasoning:<br>print(reasoning, end="")<br>if chunk.choices[0].delta.content is not None:<br>print(chunk.choices[0].delta.content, end="")

Partner EndpointsSelf-Hosted Deployments

Specifications<br>Fastest 30B A3B MoE model with leading domain accuracy for specialized agentic tasks<br>*]:min-w-0 [&>*]:shrink">Customization<br>*]:min-w-0 [&>*]:shrink">Long-running agents<br>*]:min-w-0 [&>*]:shrink">Open<br>*]:min-w-0 [&>*]:shrink">Text-to-Text

ProviderNVIDIALast Modified23 hours agoContext Length1MParameters30BInput ModalitiesTextOutput ModalitiesText

Model Availability<br>Free EndpointAvailablePartner EndpointAvailableDownload AvailableAvailable