Browsers

Cloud-hosted Chrome for browser automation

Connect Playwright or Puppeteer to Chrome running in the cloud. Save browser profiles between sessions, watch runs live, and retrieve recordings and downloads through the API.

Start for free Read the docs
Chrome
Example workflow: browsing GitHub repositories

Automate forms and signed-in pages

Reuse a login

Save cookies and browser storage in a profile, then load it in later sessions. Use separate profiles for different accounts when accessing dashboards, reports, or customer portals.

Save a browser profile

Run your browser scripts

Navigate, fill forms, and read content rendered by JavaScript. Connect through Chrome DevTools Protocol to use your automation library's locators, assertions, and page methods.

Connect to Chrome

Let a user take over

Embed Live View so users can watch the browser or control it. They can enter a verification code or complete a form, then your script can continue in the same session.

Embed Live View

Download a report from a customer portal

Create a session with the account's saved profile and open the reporting page. Your script selects a date range, waits for the report, and clicks export. If the site asks for a verification code, the account owner can enter it through Live View before the script continues.

Retrieve the downloaded file through the API, then close the session. If the export fails, inspect the recording to see where the script stopped. To save updated cookies and browser storage, enable persistChanges before running the session.

Retrieve downloaded files

Browser controls and session management

Stealth mode

Enable anti-detection features per session. Enterprise plans also include ultra stealth mode.

Residential proxies

Choose a proxy location by country, state, or city across 100+ countries. Change locations during a session or reserve a static IP. Enterprise customers can use their own proxy servers.

CAPTCHA solving

Enable CAPTCHA detection and solving on paid plans. Turn it on or off during a session.

Live view

Open, share, or embed the browser view. Choose whether users can only watch or also control the session.

Recordings

Web recordings are enabled by default. Enable video recording when you need a downloadable MP4.

Profiles

Save cookies, local storage, and session storage for later sessions. Fork a profile to create an independent copy.

Extensions

Upload your Chrome extensions once and load them into any session.

Ad and tracker blocking

Block ads, trackers, cookie prompts, and popups. Configure automatic cookie banner acceptance.

Computer actions

Click, type, drag, scroll, and take screenshots using screen coordinates, including in canvas apps and overlays.

WebMCP

List and call WebMCP tools exposed by a website across tabs and frames. Inspect tool schemas, check long-running calls, and handle forms awaiting human input.

File downloads

Retrieve files downloaded by the browser through the API, including reports and exports generated during a session.

Create and control a browser session

  1. Start a session

    Configure stealth, proxy location, profile, screen size, and timeout. The response includes a connection endpoint and live view URL.

  2. Connect an automation library

    Connect Playwright, Puppeteer, or Selenium to the session endpoint.

  3. Run your automation

    Navigate, fill forms, or extract data using your automation library. Follow the run in Live View.

  4. Close the session

    Call sessions.stop in a finally block so the session closes even if your script fails. Stopping also saves profile changes when persistChanges is enabled.

Set HYPERBROWSER_API_KEY in your environment. The Node and Python examples read the top Hacker News headline with Playwright, then close the session. The cURL example creates a session and returns its connection details.

1// yarn add @hyperbrowser/sdk playwright-core2import { Hyperbrowser } from "@hyperbrowser/sdk";3import { chromium } from "playwright-core";45const client = new Hyperbrowser({6  apiKey: process.env.HYPERBROWSER_API_KEY,7});89const session = await client.sessions.create({ timeoutMinutes: 5 });10console.log("Live view:", session.liveUrl);1112try {13  const browser = await chromium.connectOverCDP(session.wsEndpoint);14  const page = await browser.contexts()[0].newPage();15  await page.goto("https://news.ycombinator.com/");16  console.log(await page.locator(".titleline > a").first().innerText());17} finally {18  await client.sessions.stop(session.id);19}
Setup and API reference

Browser pricing

1 credit = $0.001. Proxies and CAPTCHA solving require a paid plan.

UsageCreditsPrice
Browser usage100 credits / hour$0.10 / hour
Proxy data10,000 credits / GB$10 / GB
All plans and rates

Common questions

Can I use the browser automation code I already have?

Yes. Create a session and connect Playwright or Puppeteer to the returned wsEndpoint. Use the same page methods and locators you use locally. Selenium is also supported, with a separate connection setup described in the docs.

Read more

Does a new session remember my login?

Attach a profile to reuse its cookies and browser storage. To save changes from the current session, set persistChanges to true and stop the session when finished. Without that option, the saved profile stays unchanged. A session without a profile starts with fresh browser state.

Read more

Does WebMCP work on every website?

Only sites that register WebMCP tools expose them. Enable WebMCP when creating the session, navigate to the site, and list its tools. Your agent can inspect their schemas and call them. Use ordinary browser automation on pages without WebMCP tools.

Read more

When should I use the Web API instead?

Use the Web API to fetch page content, search the web, or crawl a site. Use a browser session when your code needs to click through several pages, manage tabs, or keep a browser open between operations.

Read more

Other products

Start with a free account.

Includes 5,000 credits for browsers, sandboxes, and web data.