Nvidia Nemotron 3 Ultra 550B Thinking API
About
Build advanced reasoning agents with Nvidia Nemotron 3 Ultra 550B Thinking, a specialized AI model optimized for complex logic and chain-of-thought tasks. This model delivers cost-effective pricing at $0.50/1M input and $2.50/1M output tokens, native tool calling support, open weights architecture. Access Nvidia Nemotron 3 Ultra 550B Thinking via the NanoGPT API with up to 66K output tokens.
Capabilities
Input Modalities
text
Output Modalities
text
Reasoning
Yes
Structured Output
No
Tool Use
Yes
WeightsOpen
Limits
Context Window
1.0M
Tokens
Input Limit
1.0M
Tokens
Max Output
66K
Tokens
Pricing
Standard (per 1M tokens)
Input
$0.50
Output
$2.50
Caching (per 1M tokens)
Read
$0.25
Frequently Asked Questions
Nvidia Nemotron 3 Ultra 550B Thinking by NanoGPT costs $0.50 per 1M input tokens and $2.50 per 1M output tokens. Cached reads cost $0.25 per 1M tokens.
Model Details
ID
nvidia/nemotron-3-ultra-550b-a55b:thinking
Provider
NanoGPTFamily
nemotron
Release Date
Jun 4, 2026
Knowledge Cutoff
N/A
API Integration
NPM Package
@ai-sdk/openai-compatible
Environment Variables
NANO_GPT_API_KEY
API Base URL
https://nano-gpt.com/api/v1