# Nemotron 3 Ultra 550B

> NVIDIA's strongest US open-weight. 550B MoE (55B active), hybrid Mamba-Transformer. 1M context, 300+ tok/s.

Human version: https://kymaapi.com/models/nemotron-3-ultra-550b
Live JSON: `GET https://kymaapi.com/v1/models` (no auth required)

## Facts

- **Model ID**: `nemotron-3-ultra-550b` — pass this as `model` in the request body
- **Creator**: NVIDIA
- **Released**: 2026-06-04
- **Context window**: 1M tokens
- **Max output**: 33K tokens per response — a hard ceiling, not a default
- **Price**: $0.675 in / $3.375 out per 1M, cached input 10%
- **Capabilities**: tools, reasoning, caching

## Call it

```bash
curl https://kymaapi.com/v1/chat/completions \
  -H "Authorization: Bearer $KYMA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "nemotron-3-ultra-550b", "messages": [{"role": "user", "content": "Hello"}]}'
```

## Positioning

NVIDIA's open-weight model optimized for reasoning, coding, and long-context text generation. Engineers should use it when they need a balance of throughput and frontier-level performance without proprietary API constraints.

## About

Nemotron 3 Ultra 550B is a 550B parameter mixture-of-experts architecture with 55B active parameters per forward pass. It combines a hybrid Mamba-Transformer design to handle reasoning, coding, and general-purpose text tasks efficiently.

On Kyma, the model runs through an OpenAI-compatible endpoint with automatic request failover and exact cost reporting in the response payload. It supports tool use, structured outputs, and prompt caching, which bills repeated prefixes at 10% of the standard input rate.

The model accepts a 1,000,000 token context window but caps generation at 32,768 output tokens. It is strictly text-only and does not support vision or multimodal inputs.

## Use cases

- **Long Context Document Analysis** — Process large technical manuals or codebases within the one million token window.
- **Automated Code Generation** — Generate and refactor code using built-in reasoning and tool-use capabilities.
- **High Throughput Chat** — Handle conversational workloads that require fast text generation and structured JSON outputs.
- **Reasoning Heavy Workflows** — Execute multi-step logical tasks where active parameter routing improves accuracy.

## Not ideal for

Tasks requiring image or video understanding, or workloads that need output sequences longer than 32,768 tokens.

## Pick something else when

- You need vision or multimodal inputs → `gemini-3.5-flash`, `gpt-5.6-terra`
- You require output sequences longer than 32k tokens → `qwen-3.7-max`, `deepseek-v4-pro`
- You need maximum raw speed for simple completions → `qwen3.7-flash`, `gemini-3.6-flash`

## See also

- All models: https://kymaapi.com/models.md
- Pricing: https://kymaapi.com/pricing.md
- Other models by NVIDIA: https://kymaapi.com/models?q=NVIDIA