en · ru

Available in Russia, the CIS, and worldwide

Cloud browser infrastructure for AI agents

24/7 access to a ready browser pool for parallel work with web resources. Fits any LLM agents. Works via API without complex setup.

xClouds cloud browsers — several browser windows running in the cloud

Runs on cloud and local models

GPT
Claude
Gemini
Mistral
MiniMax
Ollama
Hugging Face
DeepSeek

Use hundreds of our cloud browsers to collect and structure web data

Why cloud browsers are better than local ones

They speed up AI agents by working with thousands of web pages at the same time. They are also more convenient, safer, and save your hardware resources.

Proxy support

You can use your own proxies or order proxies from us. Our proxies support automatic rotation and country selection.

Scale easily

An AI agent can launch dozens or hundreds of browser copies at the same time, or different versions of the same browser. All modern browsers are available.

Save RAM and CPU

Your hardware will not spend valuable memory and compute resources on cloud browser work, because the browsers run on a remote server.

Safer

All browsers run in an isolated environment on remote servers, which adds privacy and security while working with web resources.

More reliable

They handle site defenses better, stay more stable during intensive data collection, and one failed browser does not break the workflow because the rest keep running.

Run from anywhere

You or your agents can connect to cloud browsers from any device: a laptop, desktop, CI/CD, or servers.

Full compatibility with Playwright, Puppeteer, Selenium, and CDP

If you already have scripts written for these frameworks, you will not need to rewrite them. Only the connection endpoint changes; the rest of the code stays the same. After connecting, you can run dozens or hundreds of cloud browsers in parallel.

                            
import { chromium } from 'playwright';

const browser = await chromium.connect({
    wsEndpoint: 'wss://playwright.xclouds.dev/v1.58/?api_key=API_KEY'
});

                            
import puppeteer from 'puppeteer';

const browser = await puppeteer.connect({
    browserWSEndpoint: 'wss://cdp.xclouds.dev/cdp/?api_key=API_KEY'
});

                            
from selenium import webdriver

driver = webdriver.Remote(
    command_executor="https://selenium.xclouds.dev/wd/hub?api_key=API_KEY"
)

                            
import puppeteer from 'puppeteer';

const browser = await puppeteer.connect({
    browserWSEndpoint: 'wss://cdp.xclouds.dev/cdp/?api_key=API_KEY'
});

More examples at docs.xclouds.dev

Connect our browser pool to your agents in CrewAI, LangChain, and OpenClaw

import asyncio
from browser_use import Agent, Browser, ChatBrowserUse

XCLOUDS_CDP_API_KEY = "your xclouds api key"
BROWSER_USE_API_KEY = "your browser-use api key"

async def main():
    browser = Browser(cdp_url=f"wss://cdp.xclouds.dev/cdp/?api_key={XCLOUDS_CDP_API_KEY}")
    agent = Agent(
        task="Visit https://habr.com/ and search for vpn",
        browser=browser,
        llm=ChatBrowserUse(api_key=BROWSER_USE_API_KEY),
    )
    await agent.run()

asyncio.run(main())

import asyncio
from langchain_openai import ChatOpenAI
from browser_use import Agent, Browser

XCLOUDS_CDP_URL = "wss://cdp.xclouds.dev/cdp/?api_key=YOUR_API_KEY"

agent = Agent(
    task="Find the latest AI news",
    llm=ChatOpenAI(model="gpt-4o"),
    browser=Browser(cdp_url=XCLOUDS_CDP_URL),
)
asyncio.run(agent.run())

from crewai.tools import tool

XCLOUDS_SELENIUM_URL = "https://selenium.xclouds.dev/wd/hub?api_key=YOUR_API_KEY"

@tool
def selenium_browser(uri):
    """Fetch a page through the xClouds remote Selenium browser."""
    from selenium import webdriver
    from selenium.webdriver.chrome.options import Options as ChromeOptions

    driver = webdriver.Remote(
        command_executor=XCLOUDS_SELENIUM_URL,
        options=ChromeOptions(),
    )
    driver.get(uri)
    page_source = driver.page_source
    driver.quit()
    return page_source

# Set PLAYWRIGHT_SERVER_URL in .env:
# wss://playwright.xclouds.dev/v1.58/?api_key=YOUR_API_KEY

from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.connect(
        ws_endpoint="wss://playwright.xclouds.dev/v1.58/?api_key=YOUR_API_KEY"
    )

Easy to pay

We accept bank cards from Russia and the CIS

You will not have payment problems from Russia. If needed, we can issue an invoice for companies and sole proprietors.

Try first, pay later

Right after registration, you get access to a free demo cluster. We do not ask you to attach a card.

1000 credits and 60 minutes monthly

Free resources are available to you every month. Spend them on personal and work projects.

Did the solution not fit? We will refund you!

If for any reason you change your mind about using our platform, we will refund the remaining balance without trouble.

One automation endpoint connected to several browser protocols

01 How do I start using it?

Sign in with your GitHub or Yandex account. API keys and demo connection examples will be available in your dashboard. You will automatically get access to the free demo cluster. We do not ask you to attach a card while using the demo cluster.

02 How many parallel browsers can I launch?

On the demo cluster and free accounts, you can launch no more than 2 browsers at the same time. Paid accounts have no limit.

03 Will these be isolated browsers?

Yes, these are full-featured isolated browsers within the session context.

planned for 2026

We constantly develop our infrastructure

Browser profiles

You will be able to use your own browser profiles for recurring testing tasks and data collection.

Anti-bot system

We will improve our algorithms and solutions to increase your agents' stability while collecting data from any web resources.

Session debugging

We plan to add session logging so it is easier to understand failure points and unstable agent behavior.

Do you use real browsers?

Yes, we have a pool of full-featured browsers that support JavaScript rendering, header configuration, cookies, and debugging through Chrome DevTools Protocol. We also have documentation with examples for using these capabilities.

Which browsers are available?

Chrome, Chromium, Firefox, Edge, and WebKit. On request and for an additional fee, we provide access to Safari.

Can I take screenshots and PDFs?

Yes. You can take screenshots in different resolutions and save web pages as PDF documents.

Will my agent be able to sign in to the site I need and perform multi-step scenarios?

Yes. Your agent will be able to send and receive cookies, and manage headers.

Our friends

VKontakte
Yandex

Do not need it right now? Save resources for later

Every registered user receives +1000 credits and +60 minutes on the balance every month. Sign up and save your resources, then use them when you need them!