All Models / nvidia / nvidia/nemotron-3-ultra-550b-a55b

NVIDIA: Nemotron 3 Ultra API Pricing & Cost Calculator

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...

Tool CallingReasoningPrompt Caching
Input Price / 1M Tokens
$0.500
Output Price / 1M Tokens
$2.20
Prompt Cache Read / 1M
$0.100
Context Window
262K
Max out: 16K

Interactive NVIDIA: Nemotron 3 Ultra Monthly Cost Calculator

Est. Monthly (With Cache)
—
Est. Monthly (Uncached)
—
Est. Daily Run Rate
—

Volume Cost Benchmarks (75% Input / 25% Output)

Monthly Token VolumeUncached API CostWith 50% Prompt Caching
1M Tokens (Small App)$0.93$0.78
10M Tokens (Growing SaaS)$9.25$7.75
100M Tokens (Scale / Agents)$92.50$77.50
1B Tokens (Enterprise)$925.00$775.00

Deploy or Switch Providers

Best for H100 / A100 / 4090

RunPod Cloud GPUs

Spin up on-demand H100, A100, L40S, and RTX 4090 pods or serverless vLLM endpoints in seconds.

Launch GPU on RunPod →
Predictable Cloud Infra

DigitalOcean GPU Droplets

Deploy AI inference APIs, vector databases, and GPU droplets with simple monthly billing.

Try DigitalOcean →
1 API Key · 470+ Models

OpenRouter Unified API

Switch between Gemini, Claude, GPT, DeepSeek, and Llama with one OpenAI-compatible key.

Get Unified API Key →

Lower-Cost Alternatives to NVIDIA: Nemotron 3 Ultra