# NVIDIA: Nemotron 3 Ultra

- Model id: `nvidia/nemotron-3-ultra-550b-a55b`
- Served by: -
- List price per 1M tokens, input / output, from: -, with 0% platform fee. A request is billed at the list price of the host that served it.
- Context: 262K, max output 236K
- Prices read 2026-11-15 09:33 UTC
- Page: https://plaingateway.com/models/nvidia/nemotron-3-ultra-550b-a55b

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...

## Hosts

| Host | Endpoint | In / out /M | Cache read /M | Context | Served |
|---|---|---|---|---|---|
| DeepInfra | `deepinfra/fp4/flex` | $0.40 / $1.76 | $0.08 | 262K | no |
| DeepInfra | `deepinfra/fp4` | $0.50 / $2.20 | $0.10 | 262K | no |
| Kt | `kt` | $0.50 / $2.20 | $0.10 | 262K | no |
| Baseten | `baseten/fp4/us` | $0.60 / $2.40 | $0.12 | 203K | no |
| Venice | `venice/fp8` | $0.625 / $3.13 | $0.188 | 256K | no |
| DeepInfra | `deepinfra/fp4/priority` | $0.75 / $3.30 | $0.15 | 262K | no |

## First request

```sh
curl https://api.plaingateway.com/api/v1/chat/completions \
  -H "Authorization: Bearer $PLAINGATEWAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "nvidia/nemotron-3-ultra-550b-a55b", "messages": [{"role": "user", "content": "Say hello in five words."}]}'
```

Sign up at https://plaingateway.com/signup (no card), create a key, send a test call, and add credits when you want more. New accounts may get a small starting credit. The whole bootstrap for agents: https://plaingateway.com/llms.txt