Quickstart

Create an API key

In the dashboard, open your project (a new team starts with one called default) and go to API Keys. Name the key after the service that will hold it, such as inference-bureau-api, pick an environment, and choose Create key. Owners and admins can create keys.

Keys belong to one project and environment, and are shown once, so copy it into IB_API_KEY now. A key in production starts with ib_live_; one in development starts with ib_test_. Both are billed the same way. Send test request under the new key makes one short chat completion through the public API, billed like any other, to check the whole path. See Authentication.

Change the base URL

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.inferencebureau.com/v1",
  apiKey: process.env.IB_API_KEY,
});

const res = await client.chat.completions.create({
  model: "openai/gpt-4o-mini",
  messages: [{ role: "user", content: "Hello" }],
});

Or send the first request before writing any code: the project's Playground runs a call as one of its keys and shows the answer with its receipt beside it, and copies the call as curl, the OpenAI or Anthropic client, or the AI SDK.

Find the request in Logs

Every response carries an X-Request-ID header. In the project, press F and paste it to open the request's receipt in Logs, or fetch it with GET /v1/requests/:id.

Pick a model

Only certified models answer. GET /v1/models lists the ones your key may call, with context window and capabilities such as tools, structured output, and web search. Models shows the full catalog with per-provider prices.