Run concurrent browser sessions

Launch multiple browser sessions in parallel to speed up large-scale scraping, testing, or automation. Each session runs independently on Browserless infrastructure.

Prerequisites

• A Browserless API token from your account dashboard
• A Browserless plan that supports the concurrency level you need

Steps

AI Agent

Use the Browserless MCP server to run concurrent browser sessions from any MCP-compatible AI agent (Claude Desktop, Cursor, Windsurf, ChatGPT, etc.).

One prompt that installs the Browserless MCP server and runs this example. Paste it into your AI agent.

1. Connect the MCP server

Send this prompt to your AI agent to install the Browserless MCP server:

Go to https://github.com/browserless/browserless-mcp/blob/main/install.md
and follow the instructions to install the Browserless MCP server
for my client.

2. Scrape multiple URLs

Use browserless_smartscraper. Each URL can be scraped independently in parallel calls.

Use the browserless_smartscraper tool to scrape these 5 URLs
in parallel and return the content of each as markdown:
- https://scraping-sandbox.netlify.app/products
- https://scraping-sandbox.netlify.app/contact-us
- https://scraping-sandbox.netlify.app/receipt
- https://scraping-sandbox.netlify.app/dashboard
- https://scraping-sandbox.netlify.app/helix

REST API

Each REST request opens an independent browser session on Browserless. Fire them concurrently using your language's async primitives. No WebSocket connection needed.

cURL

View Full Code on GitHub

1. Send requests in parallel

Each REST API call opens an independent browser session. Run them concurrently with &:

curl -s -X POST \
  "https://production-sfo.browserless.io/screenshot?token=YOUR_API_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{ "url": "https://scraping-sandbox.netlify.app/products", "options": { "type": "png", "fullPage": true } }' \
  --output page-1.png &

curl -s -X POST \
  "https://production-sfo.browserless.io/screenshot?token=YOUR_API_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{ "url": "https://scraping-sandbox.netlify.app/contact-us", "options": { "type": "png", "fullPage": true } }' \
  --output page-2.png &

curl -s -X POST \
  "https://production-sfo.browserless.io/screenshot?token=YOUR_API_TOKEN_HERE" \
  -H "Content-Type: application/json" \
  -d '{ "url": "https://scraping-sandbox.netlify.app/receipt", "options": { "type": "png", "fullPage": true } }' \
  --output page-3.png &

wait
echo "All screenshots saved"

2. Check the output

All three screenshots are captured simultaneously. Total time is roughly the duration of the slowest single request.

All screenshots saved

Three files appear in the current directory: page-1.png, page-2.png, page-3.png.

Node.js

View Full Code on GitHub

1. Fire concurrent requests with Promise.all

import fs from 'fs';

const TOKEN = 'YOUR_API_TOKEN_HERE';
const URLS = [
  'https://scraping-sandbox.netlify.app/products',
  'https://scraping-sandbox.netlify.app/contact-us',
  'https://scraping-sandbox.netlify.app/receipt',
  'https://scraping-sandbox.netlify.app/dashboard',
  'https://scraping-sandbox.netlify.app/helix',
];

await Promise.all(
  URLS.map((url, i) =>
    fetch(
      `https://production-sfo.browserless.io/screenshot?token=${TOKEN}`,
      {
        method: 'POST',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify({ url, options: { type: 'png', fullPage: true } }),
      }
    ).then(async (res) => {
      const buf = Buffer.from(await res.arrayBuffer());
      fs.writeFileSync(`screenshot-${i + 1}.png`, buf);
      console.log(`Saved screenshot-${i + 1}.png`);
    })
  )
);

console.log('All done');

2. Check the output

Run with node concurrent.mjs. All 5 screenshots are captured in parallel. Total time is the slowest single page, not 5× sequential.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

Python

View Full Code on GitHub

1. Install dependencies

pip install requests

2. Use ThreadPoolExecutor for parallel requests

import requests
from concurrent.futures import ThreadPoolExecutor, as_completed

TOKEN = 'YOUR_API_TOKEN_HERE'
URLS = [
    'https://scraping-sandbox.netlify.app/products',
    'https://scraping-sandbox.netlify.app/contact-us',
    'https://scraping-sandbox.netlify.app/receipt',
    'https://scraping-sandbox.netlify.app/dashboard',
    'https://scraping-sandbox.netlify.app/helix',
]

def capture(i_url):
    i, url = i_url
    res = requests.post(
        f'https://production-sfo.browserless.io/screenshot?token={TOKEN}',
        json={'url': url, 'options': {'type': 'png', 'fullPage': True}},
    )
    filename = f'screenshot-{i + 1}.png'
    with open(filename, 'wb') as f:
        f.write(res.content)
    print(f'Saved {filename}')

with ThreadPoolExecutor(max_workers=5) as executor:
    futures = [executor.submit(capture, (i, url)) for i, url in enumerate(URLS)]
    for f in as_completed(futures):
        f.result()

print('All done')

3. Check the output

Run with python concurrent.py. All 5 screenshots are captured in parallel.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

Java

View Full Code on GitHub

1. Dependencies

java.net.http.HttpClient ships with the JDK (Java 11+). No extra dependencies needed.

2. Fire concurrent requests with CompletableFuture

import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.List;
import java.util.concurrent.CompletableFuture;
import java.util.stream.IntStream;

public class ConcurrentScreenshots {
    public static void main(String[] args) {
        String token = "YOUR_API_TOKEN_HERE";
        List<String> urls = List.of(
            "https://scraping-sandbox.netlify.app/products",
            "https://scraping-sandbox.netlify.app/contact-us",
            "https://scraping-sandbox.netlify.app/receipt",
            "https://scraping-sandbox.netlify.app/dashboard",
            "https://scraping-sandbox.netlify.app/helix"
        );

        HttpClient client = HttpClient.newHttpClient();

        CompletableFuture<?>[] futures = IntStream.range(0, urls.size())
            .mapToObj(i -> {
                String body = "{\"url\":\"" + urls.get(i) + "\",\"options\":{\"type\":\"png\",\"fullPage\":true}}";
                HttpRequest request = HttpRequest.newBuilder()
                    .uri(URI.create(
                        "https://production-sfo.browserless.io/screenshot?token=" + token))
                    .header("Content-Type", "application/json")
                    .POST(HttpRequest.BodyPublishers.ofString(body))
                    .build();

                return client
                    .sendAsync(request, HttpResponse.BodyHandlers.ofByteArray())
                    .thenAccept(res -> {
                        try {
                            Files.write(Path.of("screenshot-" + (i + 1) + ".png"), res.body());
                            System.out.println("Saved screenshot-" + (i + 1) + ".png");
                        } catch (Exception e) {
                            throw new RuntimeException(e);
                        }
                    });
            })
            .toArray(CompletableFuture[]::new);

        CompletableFuture.allOf(futures).join();
        System.out.println("All done");
    }
}

3. Check the output

Compile with javac ConcurrentScreenshots.java and run with java ConcurrentScreenshots. allOf blocks until the slowest request finishes.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

This example builds the JSON body via string concatenation. For production use, add a library like Jackson to serialize the body safely. URLs containing quotes or backslashes will break the raw string approach.

C#

View Full Code on GitHub

1. Dependencies

System.Net.Http.HttpClient is part of the .NET standard library. No packages needed.

2. Fire concurrent requests with Task.WhenAll

using System;
using System.IO;
using System.Linq;
using System.Net.Http;
using System.Text;
using System.Threading.Tasks;

class ConcurrentScreenshots
{
    static async Task Main()
    {
        const string token = "YOUR_API_TOKEN_HERE";
        string[] urls = {
            "https://scraping-sandbox.netlify.app/products",
            "https://scraping-sandbox.netlify.app/contact-us",
            "https://scraping-sandbox.netlify.app/receipt",
            "https://scraping-sandbox.netlify.app/dashboard",
            "https://scraping-sandbox.netlify.app/helix",
        };

        using var client = new HttpClient();

        await Task.WhenAll(urls.Select(async (url, i) =>
        {
            var body = new StringContent(
                $"{{\"url\":\"{url}\",\"options\":{{\"type\":\"png\",\"fullPage\":true}}}}",
                Encoding.UTF8,
                "application/json"
            );
            var response = await client.PostAsync(
                $"https://production-sfo.browserless.io/screenshot?token={token}",
                body
            );
            var bytes = await response.Content.ReadAsByteArrayAsync();
            await File.WriteAllBytesAsync($"screenshot-{i + 1}.png", bytes);
            Console.WriteLine($"Saved screenshot-{i + 1}.png");
        }));

        Console.WriteLine("All done");
    }
}

3. Check the output

Run with dotnet run. Task.WhenAll fires all requests concurrently and waits for the last one to finish.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

This example builds the JSON body via string interpolation. For production use, use System.Text.Json.JsonSerializer.Serialize instead. URLs containing quotes or backslashes will break the raw string approach.

PHP

View Full Code on GitHub

1. Install dependencies

composer require react/http

2. Send requests concurrently with ReactPHP

ReactPHP's event loop runs all requests in parallel without spawning threads:

<?php

require 'vendor/autoload.php';

use React\Http\Browser;
use React\EventLoop\Loop;

$token = 'YOUR_API_TOKEN_HERE';
$urls = [
    'https://scraping-sandbox.netlify.app/products',
    'https://scraping-sandbox.netlify.app/contact-us',
    'https://scraping-sandbox.netlify.app/receipt',
    'https://scraping-sandbox.netlify.app/dashboard',
    'https://scraping-sandbox.netlify.app/helix',
];

$browser = new Browser();
$promises = [];

foreach ($urls as $i => $url) {
    $index = $i + 1;
    $promises[] = $browser
        ->post(
            "https://production-sfo.browserless.io/screenshot?token={$token}",
            ['Content-Type' => 'application/json'],
            json_encode(['url' => $url, 'options' => ['type' => 'png', 'fullPage' => true]])
        )
        ->then(function ($response) use ($index) {
            file_put_contents("screenshot-{$index}.png", (string) $response->getBody());
            echo "Saved screenshot-{$index}.png\n";
        });
}

\React\Promise\all($promises)->then(function () {
    echo "All done\n";
});

Loop::run();

3. Check the output

Run with php concurrent.php. Loop::run() processes all requests through the event loop until the last promise resolves.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

Ruby

View Full Code on GitHub

1. Dependencies

net/http and json are part of Ruby's standard library. No gems required.

2. Spawn one thread per URL

require 'net/http'
require 'json'
require 'uri'

TOKEN = 'YOUR_API_TOKEN_HERE'
URLS = [
  'https://scraping-sandbox.netlify.app/products',
  'https://scraping-sandbox.netlify.app/contact-us',
  'https://scraping-sandbox.netlify.app/receipt',
  'https://scraping-sandbox.netlify.app/dashboard',
  'https://scraping-sandbox.netlify.app/helix',
].freeze

threads = URLS.each_with_index.map do |url, i|
  Thread.new do
    uri = URI("https://production-sfo.browserless.io/screenshot?token=#{TOKEN}")
    http = Net::HTTP.new(uri.host, uri.port)
    http.use_ssl = true

    request = Net::HTTP::Post.new(uri)
    request['Content-Type'] = 'application/json'
    request.body = JSON.generate({ url: url, options: { type: 'png', fullPage: true } })

    response = http.request(request)
    File.binwrite("screenshot-#{i + 1}.png", response.body)
    puts "Saved screenshot-#{i + 1}.png"
  end
end

threads.each(&:join)
puts 'All done'

3. Check the output

Run with ruby concurrent.rb. join waits for all threads before printing the final line.

Saved screenshot-1.png
Saved screenshot-2.png
Saved screenshot-3.png
Saved screenshot-4.png
Saved screenshot-5.png
All done

Frameworks

Each connect() call opens a separate managed browser on Browserless. Use Promise.all or goroutines to run them in parallel. Sessions are completely isolated from each other.

Puppeteer

View Full Code on GitHub

1. Install dependencies

npm install puppeteer-core

2. Open multiple browser connections in parallel

Each puppeteer.connect() call opens a separate browser session on Browserless:

import puppeteer from 'puppeteer-core';

const TOKEN = 'YOUR_API_TOKEN_HERE';
const WS = `wss://production-sfo.browserless.io?token=${TOKEN}`;

const URLS = [
  'https://scraping-sandbox.netlify.app/products',
  'https://scraping-sandbox.netlify.app/contact-us',
  'https://scraping-sandbox.netlify.app/receipt',
  'https://scraping-sandbox.netlify.app/dashboard',
  'https://scraping-sandbox.netlify.app/helix',
];

const results = await Promise.all(
  URLS.map(async (url, i) => {
    const browser = await puppeteer.connect({ browserWSEndpoint: WS });
    try {
      const page = await browser.newPage();
      await page.goto(url, { waitUntil: 'networkidle2' });
      const title = await page.title();
      console.log(`[${i + 1}] ${title}`);
      return { url, title };
    } finally {
      // Always close to release the session even on error.
      await browser.close();
    }
  })
);

console.log('Results:', results);

3. Check the output

Run with node concurrent.mjs. Each session runs in its own browser process on Browserless. There's no shared state between them.

[1] Product Catalog | Scraping Sandbox
[2] Contact Us - Scraping Sandbox | Scraping Sandbox
[3] Scraping Sandbox
[4] Scraping Sandbox
[5] Careers at Helix — Open Positions
Results: [ { url: '...', title: '...' }, ... ]

Playwright

View Full Code on GitHub

1. Install dependencies

npm install playwright-core

2. Open multiple connections in parallel

import { chromium } from 'playwright-core';

const TOKEN = 'YOUR_API_TOKEN_HERE';
const WS = `wss://production-sfo.browserless.io?token=${TOKEN}`;

const URLS = [
  'https://scraping-sandbox.netlify.app/products',
  'https://scraping-sandbox.netlify.app/contact-us',
  'https://scraping-sandbox.netlify.app/receipt',
  'https://scraping-sandbox.netlify.app/dashboard',
  'https://scraping-sandbox.netlify.app/helix',
];

const results = await Promise.all(
  URLS.map(async (url, i) => {
    const browser = await chromium.connectOverCDP(WS);
    try {
      // Use the default context — browser.newPage() creates a new context that
      // doesn't inherit proxy, profile, or launch settings.
      const context = browser.contexts()[0];
      const page = await context.newPage();
      await page.goto(url, { waitUntil: 'networkidle' });
      const title = await page.title();
      console.log(`[${i + 1}] ${title}`);
      return { url, title };
    } finally {
      // Always close to release the session even on error.
      await browser.close();
    }
  })
);

console.log('Results:', results);

3. Check the output

Run with node concurrent.mjs. Each session runs in its own browser process on Browserless.

[1] Product Catalog | Scraping Sandbox
[2] Contact Us - Scraping Sandbox | Scraping Sandbox
[3] Scraping Sandbox
[4] Scraping Sandbox
[5] Careers at Helix — Open Positions
Results: [ { url: '...', title: '...' }, ... ]

Go (chromedp)

View Full Code on GitHub

1. Install dependencies

go get github.com/chromedp/chromedp

2. Spawn one goroutine per URL

Each goroutine gets its own NewRemoteAllocator context, which opens a separate browser session on Browserless:

package main

import (
    "context"
    "fmt"
    "log"
    "sync"

    "github.com/chromedp/chromedp"
)

func main() {
    token := "YOUR_API_TOKEN_HERE"
    ws := fmt.Sprintf("wss://production-sfo.browserless.io?token=%s", token)

    urls := []string{
        "https://scraping-sandbox.netlify.app/products",
        "https://scraping-sandbox.netlify.app/contact-us",
        "https://scraping-sandbox.netlify.app/receipt",
        "https://scraping-sandbox.netlify.app/dashboard",
        "https://scraping-sandbox.netlify.app/helix",
    }

    var wg sync.WaitGroup

    for i, url := range urls {
        wg.Add(1)
        go func(i int, url string) {
            defer wg.Done()

            // Each goroutine gets its own allocator — a separate browser session per URL.
            allocCtx, cancel := chromedp.NewRemoteAllocator(
                context.Background(), ws, chromedp.NoModifyURL,
            )
            defer cancel()

            ctx, cancel := chromedp.NewContext(allocCtx)
            defer cancel()

            var title string
            if err := chromedp.Run(ctx,
                chromedp.Navigate(url),
                chromedp.Title(&title),
            ); err != nil {
                log.Printf("[%d] error: %v", i+1, err)
                return
            }

            fmt.Printf("[%d] %s\n", i+1, title)
        }(i, url)
    }

    wg.Wait()
    fmt.Println("All done")
}

3. Check the output

Run with go run main.go. Each goroutine runs independently. Output lines may arrive out of order since sessions finish at different times.

[3] Scraping Sandbox
[1] Product Catalog | Scraping Sandbox
[5] Careers at Helix — Open Positions
[2] Contact Us - Scraping Sandbox | Scraping Sandbox
[4] Scraping Sandbox
All done

Tips

• Concurrency limits — your plan determines how many sessions can run simultaneously. Requests beyond that limit queue automatically.
• Reuse a saved profile — add &profile=my-profile to the WebSocket URL to start every concurrent session pre-authenticated. See Save Logins to Authenticated Profiles.
• Rate limiting — add a concurrency cap in your own code (e.g. p-limit in Node.js) to avoid overwhelming the target site.

Next steps

Save Logins to Authenticated Profiles

share a login state across all concurrent sessions

Scrape Structured Data

extract data at scale from many pages in parallel

Take a Screenshot

capture screenshots in bulk