One API.
Your application.Your models.Your control.Your application.

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://www.napsix.com/api/gateway/v1",
  apiKey: process.env.NAPSIX_API_KEY,
});

// Mato v1.0 — nuestro modelo insignia: gratis en todos los planes.
const response = await client.chat.completions.create({
  model: "napsix/mato-v1.0",
  messages: [{ role: "user", content: "Hola Mato" }],
});

console.log(response.choices[0].message.content);

// Todos los modelos disponibles:
// const models = await client.models.list();

Integration example. Set NAPSIX_API_KEY and NAPSIX_MODEL_ID on your server; do not expose keys in the browser.

• 01

  A compatible interface

  Use the OpenAI-compatible chat completions format. Validate which parameters each model supports.

• 02

  Model selection

  Choose a model enabled for your task and plan from the platform catalog.

• 03

  BYOK by plan

  Connect your own keys when your plan includes it. Provider usage remains subject to its own terms.

• 04

  Usage monitoring

  Review application usage in the Gateway console and compare it with your plan limits.

• 05

  Streaming and tools

  Check the selected model's capabilities before adding streaming, tools or multimodal inputs.

• 06

  API reference

  Consult the OpenAPI schema and use the example URL to prepare your integration.