Nemotron 3 Ultra
nvidia/nemotron-3-ultra-550b-a55b
byNVIDIA203K contextReleased Jun 4, 2026ToolsReasoning
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...
- Input
- $0.652
- per 1M tokens
- Output
- $2.61
- per 1M tokens
- Cached input
- $0.13
- per 1M tokens
Lowest price across providers. You are billed from your Ditfa Platform credits per token used.
Providers
Requests are routed to the cheapest healthy provider and fall back automatically.
Usage
Tokens served per day, last 30 days
No traffic yet. Be the first to try it.
Top apps
Apps that send an HTTP-Referer or X-Title header, last 30 days
No apps have identified themselves yet.
Quickstart
Use any OpenAI SDK. Point it at the Ditfa Platform base URL and use an sk-dp- key.
curl https://gateway.ditfa.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $DITFA_API_KEY" \
-d '{
"model": "nvidia/nemotron-3-ultra-550b-a55b",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}'