CloudBrowser runs real Chrome in the cloud so ChatGPT, Claude and your own agents can browse, click, type and pull data from any website — while you watch live and take over anytime.
Screen recording of a real Chrome session: an AI agent is asked to “Add a sweater that fits my 30 lb dog to my cart”. It opens demo.vercel.store, closes a popup, searches for “dog sweater”, opens the product, reads the size options, picks 20 – 50 lbs and adds it to the cart.
Works with the AI tools you already use
The platform
Every option runs on the same isolated Chrome browsers, saved sessions and live view — pick the one that fits how you work.
A browser-sessions workspace and five browser skills, right inside ChatGPT.
One URL gives Claude, Cursor, VS Code, Codex and any MCP client 20 browser tools.
Open stealth Chrome with one HTTP call, then drive it over CDP with Puppeteer or Playwright.
Official .NET and Node.js SDKs, plus drop-in toolkits for the OpenAI Agents SDK and Claude Agent SDK.
How it works
Paste one URL into ChatGPT, Claude or your IDE — or call the API. Sign in with your CloudBrowser account or use an API token.
Describe the task in plain words. Your agent opens a cloud browser and picks the right tools to get it done.
Follow along on a live remote desktop. When a login or CAPTCHA needs a human, take over — then hand it back.
Model Context Protocol
Add CloudBrowser once and your assistant gets 20 browser tools — in chat, in your editor or inside your own agents.
https://mcp.cloudbrowser.aiSign in with your CloudBrowser account (OAuth) or send your API token as a Bearer header. Use the URL exactly as shown.
CloudBrowser with the MCP server URL.Developer mode is a ChatGPT beta for eligible paid plans on the web; in Business workspaces an admin sets it up. Menu names vary between ChatGPT releases.
The open-source MCP connector is free. Browser usage requires a CloudBrowser account and a paid plan after the 7-day Basic trial. Get your API token in the dashboard.
20 MCP tools
Your assistant decides what to call. CloudBrowser runs each step in a real browser and returns what the agent needs — page content, elements or a screenshot.
Spin up isolated Chrome instances on demand and keep track of every session.
open_browserOpen a stealth Chrome browser in the cloudconnect_to_browserTake direct control of a browserget_browsersList active browsersget_connected_browsersList browsers under controldisconnect_browserRelease control of a browserclose_browserClose a browser when the job is doneLoad any page, including JavaScript-heavy apps, and read what a person would see.
navigate_to_urlGo to any URL and wait for it to loadget_page_contentRead the title, text, headings, links and formsget_elementsExtract elements that match a CSS selectortake_screenshotCapture the viewport or the full pageClick, type and run scripts to search, filter, fill forms and move through flows.
click_elementClick buttons, links and menustype_textType into inputs and formsevaluate_scriptRun JavaScript on the pageWatch live, step in when a human is needed and reuse signed-in sessions.
start_remote_desktopOpen a live view of the browserstop_remote_desktopClose the live viewdetect_captchaSpot reCAPTCHA, hCaptcha and Turnstilebrowser_workspaceShow browsers and sessions in a workspaceget_saved_sessionsList saved sessionsresume_saved_sessionReopen a saved session, still signed inremove_saved_sessionDelete a saved sessionHuman in the loop
Every browser can stream to a live remote desktop. When an agent meets a login, a one-time code or a CAPTCHA, you step in — and it carries on from there.
detect_captcha tells the agent when to ask for helpChatGPT plugin · coming soon
Our ChatGPT plugin is in review. It adds a browser-sessions workspace beside the chat and five skills — research, extract data, act on websites, check journeys and resume signed-in sessions.
Screen recording of the CloudBrowser plugin inside ChatGPT. A saved browser session is resumed in the Browser sessions workspace and cloudbrowser.ai loads in it. A question typed in the workspace’s “Ask about this page” box is sent to ChatGPT, which reads the page through CloudBrowser and answers with a one-sentence summary. The browser is then closed and can be resumed again.
Built for production
Real sites have logins, heavy JavaScript and bot checks. CloudBrowser is made for all of it.
Every browser runs in its own isolated container — Chrome by default, or Chromium and Chrome Headless Shell when you need them.
Browser fingerprints match real-world devices — timezone, locale, WebGL and screen — to keep bot flags down.
Every browser gets an automatically assigned premium proxy on every plan — or bring your own when you need a specific network.
Save and restore cookies and localStorage so agents pick up exactly where they left off.
Run 10 concurrent instances on Basic and 25 on Premium, with up to 3 tabs each.
Connect over the Chrome DevTools Protocol with Puppeteer or Playwright, or use the official .NET and Node.js SDKs.
For developers
Open a browser with one API call, then connect over CDP with the tools you already know. Official SDKs for .NET and Node.js.
import puppeteer from 'puppeteer-core';
// 1. Open a browser in the cloud
const res = await fetch('https://production.cloudbrowser.ai/api/v1/Browser/Open', {
method: 'POST',
headers: {
Authorization: `Bearer ${process.env.CLOUDBROWSER_API_TOKEN}`,
'Content-Type': 'application/json',
},
body: JSON.stringify({ keepOpen: 300 }),
});
const { address } = await res.json();
// 2. Drive it like any local Chrome
const browser = await puppeteer.connect({ browserWSEndpoint: address });
const page = await browser.newPage();
await page.goto('https://example.com');
await page.screenshot({ path: 'example.png' });
await browser.close();Try asking
Once CloudBrowser is connected, these work as plain prompts in ChatGPT, Claude, Cursor or any MCP client.
“Open the pricing pages of our five main competitors and summarize every plan change in a table.”
navigate_to_urlget_page_content“Find dental clinics in Austin that offer online booking and collect name, website and phone number.”
navigate_to_urlget_elements“Check these three product pages and tell me which one is cheapest right now, with screenshots.”
navigate_to_urltake_screenshot“Walk through the signup flow on our staging site and screenshot each step. Flag anything that breaks.”
click_elementtype_text“Sign in to the supplier portal with my saved session and list last month’s invoices.”
get_saved_sessionsget_elements“Collect the ten latest posts from this blog and turn them into a Markdown digest.”
get_page_contentevaluate_scriptPricing
Start with a 7-day free trial on Basic. Every plan includes premium proxies and live remote desktop.
For startups and small projects.
$25/ month
Billed monthly. Annual plans include 2 months free.
30,000 usage units monthly
For growing businesses.
$90/ month
Billed monthly. Annual plans include 2 months free.
120,000 usage units monthly
For teams with a bigger picture.
Let’s talk
A plan shaped around your workload.
Contact usAll paid plans include a 14-day money-back guarantee.
Estimate your browser usage and choose a plan with confidence.
A cloud browser is a real web browser that runs on remote servers instead of your own computer. CloudBrowser starts isolated Chrome instances on demand, so AI agents and scripts can browse, click, type and pull data from any website — without you installing, patching or scaling Chrome.
Any client that supports remote MCP servers: ChatGPT, Claude, Claude Code, Cursor, VS Code, Codex, Gemini CLI, n8n and more. Add one URL — https://mcp.cloudbrowser.ai — and the assistant gets CloudBrowser’s 20 browser tools.
Yes. The CloudBrowser plugin for ChatGPT is in review and coming soon to ChatGPT’s plugin directory. It adds a browser-sessions workspace and five browser skills to ChatGPT. Until it is listed, you can connect the MCP server in ChatGPT developer mode.
The open-source MCP connector is free, and it adds no AI inference charge. Browser usage requires a CloudBrowser account and a paid plan after the 7-day Basic trial.
No. You can connect ChatGPT or Claude with a URL and manage browsers from the dashboard. Developers can use the HTTP API, Puppeteer, Playwright and our .NET and Node.js SDKs.
There are official SDKs for .NET and Node.js. Any language that can send an HTTP request and speak the Chrome DevTools Protocol works too — Python with Playwright, for example.
One usage unit is 30 seconds of browser time per tab, so two tabs open for one minute use four units. Basic includes 30,000 units a month and Premium 120,000. The usage calculator helps you estimate.
Yes. Live remote desktop is included on every plan. Open it from the dashboard or let your agent start it with start_remote_desktop, then step in whenever a human is needed.
Browsers can save their session — cookies and localStorage — under a label and restore it later, so an agent can return to a site you have already signed in to. Saved sessions do not reopen tabs.
Research and data extraction, price and content monitoring, lead lists, QA of signup and checkout flows, screenshots and back-office work in portals you have access to — anything a person can do in Chrome. You are responsible for following each website’s terms.
Start with a 7-day free trial. Connect ChatGPT, Claude or your code in minutes.
Talk to us
Custom concurrency, dedicated proxies or an agent workload you want to run by us — tell us what you are building.
contact@cloudbrowser.ai