Nvidia logo

Nemotron 3 Ultra – Benchmarks, Pricing & Intelligence Analysis

550B

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...

Not available on LLMBase

Nemotron 3 Ultra is not currently available in LLMBase Chat or the LLMBase Inference API. Benchmarks, pricing, and model details remain available for research and comparison.

Input Price$0.68/1M tokens
Output Price$2.67/1M tokens
Intelligence37.8
Coding49.3

Specifications

Technical details and pricing.

ProviderNvidia
Context Window512,288 tokens
Release DateJun 4, 2026
ModalitiesText
CapabilitiesFunction Calling
AvailabilityNot available

Benchmarks

7 benchmark scores from Artificial Analysis.

GPQA86.7%
HLE26.6%
SciCode39.9%
LCR67.0%
IFBench81.4%
Tau283.3%
TerminalBench Hard36.4%

Composite Indices

Higher is better; speed and price are normalized

Standard Benchmarks

Only benchmarks with data are shown

EU-Hosted Inference API

Power your AI projects with the best open-source models.

Drop-in OpenAI-compatible API. No data leaves Europe.

Explore Inference API

MiniMax

MiniMax M3

$0.40 / $1.40

per M tokens

Z.ai

GLM 5.2

$1.40 / $4.40

per M tokens

MoonshotAI

Kimi K2.7 Code

$0.95 / $4.00

per M tokens

DeepSeek

DeepSeek V4 Pro

$1.80 / $3.60

per M tokens

Frequently Asked Questions

What is Nemotron 3 Ultra good for?

Use Nemotron 3 Ultra for everyday tasks like writing, summarizing, brainstorming, and getting clear explanations.

How much does Nemotron 3 Ultra cost?

Pricing is based on usage. Current rates are $0.68/1M tokens for input and $2.67/1M tokens for output.

Is Nemotron 3 Ultra available in LLMBase Chat yet?

Not yet on this page. We will add chat and inference actions automatically once an exact catalog entry exists for Nemotron 3 Ultra.

Does Nemotron 3 Ultra support images or audio?

Nemotron 3 Ultra focuses on text-based tasks.

Benchmarks and pricing use Artificial Analysis where available. Catalog specs are used as a fallback.