Quick start

Three steps: give the pool an account, mint a key, and point a client at the base URL. You will have a streamed reply in a couple of minutes.

1. Add a session

   The pool needs at least one live ChatGPT session. Capture one on your own machine and it uploads straight into the gateway:

   python get_session.py --name main

   Or open the dashboard, expand Add a session, and paste a captured JSON file. See Capturing sessions for the full walkthrough and the browser extension.

2. Create an API key

   In the dashboard, open API keys and click Generate. Copy the msk-… value once — it is shown in full only at creation. Hand that key to whatever will call the gateway, and keep your admin key to yourself.

3. Point a client at the base URL

   Any OpenAI-compatible client works. Set the base URL and the key you just made:

   Setting	Value
   Base URL	__BASE__/v1
   API key	msk-… (the one you generated)
   Model	auto

Your first request

A plain curl against the gateway — swap in your key:

curl __BASE__/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"messages":[{"role":"user","content":"Hello!"}]}'

Add "stream": true to receive the answer incrementally as Server-Sent Events:

curl __BASE__/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"stream":true,"messages":[{"role":"user","content":"Write a haiku about the sea."}]}'

Using the OpenAI SDK

from openai import OpenAI

client = OpenAI(
    base_url="__BASE__/v1",
    api_key="YOUR_API_KEY",
)

resp = client.chat.completions.create(
    model="auto",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "__BASE__/v1",
  apiKey: "YOUR_API_KEY",
});

const r = await client.chat.completions.create({
  model: "auto",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(r.choices[0].message.content);