NVIDIA

NVIDIA Nemotron 3 Ultra

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

Context window
262Ktokens
Max output
183Ktokens
Input
$0.60/ 1M
Output
$2.40/ 1M

Quick start

from openai import OpenAI

client = OpenAI(
    base_url="https://neurongate.net/v1",
    api_key="sk-ng-your-api-key",
)

completion = client.chat.completions.create(
    model="nvidia/nemotron-3-ultra-550b-a55b",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(completion.choices[0].message.content)

Cost

Example requestTokensCost
Short prompt1K in$0.0006
Typical chat turn1K in · 500 out$0.0018
Document analysis10K in · 2K out$0.0108
Long context100K in · 10K out$0.084
Estimated cost per request$0.0018

Prices in USD. Billed per token actually used, from a balance prepaid in USDT, USDC, ETH or BTC.

Frequently asked questions

What is NVIDIA Nemotron 3 Ultra?

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

How much does NVIDIA Nemotron 3 Ultra cost?

NVIDIA Nemotron 3 Ultra is priced at $0.6 per 1 million input tokens and $2.4 per 1 million output tokens when accessed via NeuronGate. You pay per token — no subscriptions required. Top up with crypto (USDT, USDC, ETH, BTC).

Does NVIDIA Nemotron 3 Ultra support streaming?

Yes, NVIDIA Nemotron 3 Ultra supports streaming responses via NeuronGate's API. Use the standard "stream": true parameter in your request.

What is the context window of NVIDIA Nemotron 3 Ultra?

NVIDIA Nemotron 3 Ultra has a context window of 262K tokens (~197K words). Maximum output is 183K tokens.

Does NVIDIA Nemotron 3 Ultra support function calling (tools)?

Yes, NVIDIA Nemotron 3 Ultra supports function calling / tool use via NeuronGate's standard API.

How do I use NVIDIA Nemotron 3 Ultra with NeuronGate?

Create a NeuronGate account, top up your balance with crypto, generate an API key, and use the model id nvidia/nemotron-3-ultra-550b-a55b in your requests. NeuronGate uses the OpenAI-compatible API format — set your base URL to https://neurongate.net/v1.