RouterBase OpenAI Config

RouterBase exposes an OpenAI-compatible API at https://routerbase.com/v1. This package gives JavaScript and TypeScript projects a tiny, dependency-free way to configure OpenAI-style clients for RouterBase.

Install

npm install routerbase-openai-config

Use With The OpenAI SDK

import OpenAI from "openai";
import { createRouterBaseOpenAIOptions } from "routerbase-openai-config";

const client = new OpenAI(
  createRouterBaseOpenAIOptions({
    apiKey: process.env.ROUTERBASE_API_KEY
  })
);

const response = await client.chat.completions.create({
  model: "google/gemini-2.5-flash",
  messages: [{ role: "user", content: "Hello RouterBase" }]
});

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

Helpers

import {
  ROUTERBASE_BASE_URL,
  DEFAULT_ROUTERBASE_MODEL,
  createRouterBaseOpenAIOptions,
  createRouterBaseHeaders,
  resolveRouterBaseBaseURL
} from "routerbase-openai-config";

• ROUTERBASE_BASE_URL: https://routerbase.com/v1
• DEFAULT_ROUTERBASE_MODEL: google/gemini-2.5-flash
• createRouterBaseOpenAIOptions(options): returns { apiKey, baseURL, defaultHeaders }
• createRouterBaseHeaders(options): returns JSON and bearer-token headers for direct fetch calls
• resolveRouterBaseBaseURL(value): normalizes trailing slashes

Direct Fetch Example

import {
  ROUTERBASE_BASE_URL,
  createRouterBaseHeaders
} from "routerbase-openai-config";

const response = await fetch(`${ROUTERBASE_BASE_URL}/chat/completions`, {
  method: "POST",
  headers: createRouterBaseHeaders({
    apiKey: process.env.ROUTERBASE_API_KEY
  }),
  body: JSON.stringify({
    model: "google/gemini-2.5-flash",
    messages: [{ role: "user", content: "Say hello in five words." }]
  })
});

console.log(await response.json());

Links

• RouterBase
• RouterBase docs
• Chat completions docs