Chrome browsers in our cloud, three layers to drive them. Start with the Step API: 20 functions that navigate, click, type, scrape, solve captchas, and call AI mid-run, so you write less code. Drop to CDP for raw control, or run whole automations over REST.
Try for free, with 2 hours of runtime.
Start here
Twenty functions that each do a whole job: navigate, click, type, solve a captcha, scrape a paginated list, run an AI prompt mid-session. Call them from Node, a CDP client, or an agent over MCP.
Explore the Step APIIn Puppeteer or Playwright you write the mechanics yourself: waitForSelector, $$eval, click next, wait, loop. A step is that sequence already written. Pass it what it needs: click takes the button selector, scrape takes the data selector, the next-page selector, and a result count. A dozen lines come out as one.
Give your agents hands on a real browser. The steps become tools an MCP client calls on its own: scrape a page for the data it needs, or chain a series to log in, fill a form, and read the result back. Claude, Cursor, or an agent you built yourself, all on your account key. The axiom.ai MCP server ships in the next release.
Steps are ordinary async functions, so they can sit in the middle of the JavaScript you already write. Pull rows from your database and loop them through a step. Drop to raw CDP with Puppeteer or Playwright when a step doesn't cover something, then carry on with steps in the same session, on the same browser.
The steps run on the same engine as the No-Code Tool, so they wait for elements, retry flaky actions, and time out sensibly by default. Long-running steps poll transparently until they finish.
1import { AxiomApi } from 'axiom-api';2 3const axiom = new AxiomApi(process.env.AXIOM_API_KEY);4 5await axiom.browserOpen();6try {7 await axiom.goto("https://example.com/products");8 await axiom.click("button[data-load-more]");9 const rows = await axiom.scrape({ maxResults: 200 });10 console.log(rows.length);11} finally {12 await axiom.browserClose();13}The other two layers
The Step API sits between two layers: CDP below it for protocol-level control, Orchestrate above it for running no-code automations.
A CDP WebSocket to fresh cloud Chrome. Puppeteer, Playwright, or any language that speaks the protocol. Existing scripts run as-is.
Read more about the Chrome API
Trigger any pre-built automation, code or no-code, and manage its runs over REST. For Zapier, Make, cron, or a backend.
Read more about the Orchestrate API
Backed by
Combinator
Total bots run
Total runtime
Total users
Docs
Quickstarts, worked examples, and full endpoint references for each layer. Start with the quickstart, generate a key, and build from there.
Start with the quickstartAgents
Interested in what we're building? Contact support for early access.
Contact supportGive an agent a prompt. It authors the automation itself, with the same compiler the Builder uses.
It tests its work on the same cloud browsers and reads the same run reports you do.
Failed run, read the errors, revise, re-run. It pulls you in when a decision needs a human, not for every retry.
Build, run, observe, and fix will ship as MCP tools. Any agent with an axiom.ai API key can run the loop.