NVIDIA Nemotron 3.5 Lightning 30B

ServerlessDedicated
Hugging Face IDnvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16View on Hugging Face ↗

NVIDIA Nemotron 3.5 Lightning 30B is a compact hybrid Mamba-Transformer mixture of experts from NVIDIA. It has 30B parameters, 3B active per token, and up to 1M tokens of context.

Ways to run NVIDIA Nemotron 3.5 Lightning 30B

Dedicated endpoint

Get a quote →

Your own GPUs, tuned for your traffic. We profile your workload, tune the deployment and put throughput and latency in your contract.

Serverless

API docs →

Pay per token on shared GPUs, with no capacity to plan or reserve.

Inside your network

Security →

Run a dedicated endpoint inside your own network, including air-gapped environments.

Pricing

Dedicated endpoint
Per GPU-hour
We size the GPUs to your traffic.
Serverless
$0.0125 / $0.05
Per 1M tokens (input/output)

Use it with the agents we set up for you

Point any open-source agent at your endpoint, or pick from our catalog and we set it up as part of your contract.

Cline CodingOpenCode CodingAider Coding
GPT Researcher ResearchRAGFlow DocumentsNanobrowser Browser

NVIDIA Nemotron 3.5 Lightning 30B FAQ

How much does NVIDIA Nemotron 3.5 Lightning 30B cost on Saturn Cloud?

Serverless costs $0.0125 per 1M input tokens and $0.05 per 1M output tokens. A dedicated endpoint is priced per GPU-hour for the GPUs it runs on, and we size the GPUs to your traffic.

What is Nemotron 3.5 Lightning 30B's context length?

Up to 1M tokens. NVIDIA’s card runs it at 256K on a single H100.

Can I turn off reasoning in Nemotron 3.5 Lightning 30B?

Yes. Reasoning is on by default. Set enable_thinking to false in chat_template_kwargs to turn it off.

Does Nemotron 3.5 Lightning 30B support tool calling?

Yes. On vLLM and SGLang it uses the qwen3_coder tool-call parser. For coding agents, NVIDIA recommends force_nonempty_content.

What GPU does Nemotron 3.5 Lightning 30B need?

It runs on one H100 80GB or A100 80GB.