Overview - QX Labs Docs

The QX API lets your own code use what you build in QX: send messages to agents, execute flows and grids, and read their results. Every call runs as you, inside your org, and uses your org’s credits the way the app does.

Get your key

In the app, open Settings → API Access and request the products you need.

The API Access page in Settings, with the products to request and a box for your use case

Once your request is approved, open api.qxlabs.com, sign in with Continue with QX, and copy your key from the dashboard.

Send a request

Every endpoint lives under https://api.qxlabs.com/v1. Send your key as a bearer token in the Authorization header. The X-API-Key header works too.

curl https://api.qxlabs.com/v1/responses \
  -H "Authorization: Bearer <YOUR_QX_API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{"model": "builtin:qx_agent", "input": "Say hello in five words"}'

import requests

response = requests.post(
    "https://api.qxlabs.com/v1/responses",
    headers={"Authorization": "Bearer <YOUR_QX_API_KEY>"},
    json={"model": "builtin:qx_agent", "input": "Say hello in five words"},
)
# The agent's message is the last item in output
print(response.json()["output"][-1]["content"][0]["text"])

import axios from "axios";

const response = await axios.post(
  "https://api.qxlabs.com/v1/responses",
  { model: "builtin:qx_agent", input: "Say hello in five words" },
  { headers: { Authorization: "Bearer <YOUR_QX_API_KEY>" } },
);
// The agent's message is the last item in output
console.log(response.data.output.at(-1).content[0].text);

Every endpoint page has a Try it button that sends the request from your browser with your key, so you can see a real response before you write any code.

What you can call

Agents

Message any agent through OpenAI’s Responses API, with files, streaming and approvals.

Flows

Execute a flow with values for its input nodes and read each node’s output.

Grids

Execute a sheet, or part of it, and read its rows.

Your key covers the products an admin approved, so a call to another product’s endpoint gets a 403. The agents endpoints follow OpenAI’s Responses API, so OpenAI’s SDKs work with your QX key.

Limits and errors

Each product has its own limits on reads and writes per minute. A call over a limit gets a 429 with a message naming the limit. Flows and grids endpoints return errors as {"error": "message"}, with the app’s own message when a flow or sheet can’t execute or your org is out of credits. Agents endpoints return errors in OpenAI’s shape, with error.message, type, param and code.

​Get your key

​Send a request

​What you can call