Volver a modelos
NVIDIA·Familia Nemotron
Nemotron 3 Ultra
MoEotherPublicado el 4 jun 2026
NVIDIA's flagship open 550B-parameter Mixture-of-Experts model with 55B active parameters, built for frontier reasoning and orchestration in long-running agentic systems. Features hybrid Mamba-Transformer architecture, LatentMoE routing, multi-token prediction, and NVFP4 precision for 5x higher throughput. Achieves 30% lower cost-to-task-completion on agentic benchmarks. Supports 1M+ token context window with 95% accuracy on Ruler@1M.
ChatCompletar textoLlamada a funcionesGeneración de códigoRazonamientotext · code
Disponible en 4 proveedores
Más económico
NVIDIA NIM
Gratis / Gratis
Deep Infra, NVIDIA NIM, OpenRouter, Together AI
Proveedores (4)
Ordenados por coste total (entrada y salida por 1 millón de tokens). Selecciona una fila para ver los detalles del proveedor.
| Proveedor | Precio (por 1 M) | Límites | Regiones | Estado | Latencia |
|---|---|---|---|---|---|
| Entrada: GratisSalida: Gratis | 100 RPM / 500K TPM | us-east-1us-west-2global | Operativo | 0ms |
| Entrada: $0.40Salida: $1.60 | 200 RPM / 500K TPM | us-east-1eu-west-1 | Operativo | 0ms |
| Entrada: $0.40Salida: $1.60 | 200 RPM / 1.0M TPM | us-east-1 | Operativo | 0ms |
| Entrada: $0.50Salida: $1.50 | 600 RPM / 1.0M TPM | us-east-1us-west-2 | Operativo | 0ms |
Inicio rápido
Usa este modelo mediante NVIDIA NIM con un SDK compatible con OpenAI.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.nvidia.com/v1",
apiKey: process.env.NVIDIA_API_KEY,
});
const response = await client.chat.completions.create({
model: "nvidia/nemotron-3-ultra-550b-a55b",
messages: [
{ role: "user", content: "Hello!" }
],
});
console.log(response.choices[0].message.content);API de NVIDIA NIM · SDK compatible con OpenAI