BUILT FOR AGENTS. READY FOR THE WEB.

Browser infrastructure
for AI agents.

Give your agents a real browser.
Create sessions, persist context, and automate the web through one API.

Your agent. A real browser. One API.
INSIDE THE CONTROL PLANEPRODUCT PREVIEW / DEMO DATA
app.0browser.io
WORKSPACE / DEMO

Browser sessions

+ New session
ACTIVE SESSIONS03 running
PERSISTENT PROFILES08 available
SESSIONSTATUSPROFILE
research-agentActiveresearch-demo
qa-checkoutActiveqa-demo
data-collectorActivecollector-demo
Live viewerresearch-agent
example.com/research
THE RESEARCH COLLECTION

A little more
perspective.

Explore collection
Agent connected
browser.connect()CDP connection established
FITS RIGHT INTO YOUR STACK
PlaywrightPuppeteerChrome DevTools ProtocolREST API
01 / THE FOUNDATION

Build the agent.
We handle the browser.

From the first page load to the next thousand steps, give your automation a browser it can work with.

SESSIONS

A browser for every task.

Create isolated browser sessions on demand. Start a run, do the work, and close it through the API.

PERSISTENT PROFILES

Pick up where you left off.

Reuse persistent profiles, cookies, and browser state across runs. Keep your workflows in context.

PROXY CONFIGURATION

Choose your connection.

Configure a proxy for each session and set its timezone. Make network settings part of your workflow.

LIVE VIEWER

See what your agent sees.

Follow a session in the live viewer. Inspect the page and interact with the browser from your console.

CDP INTEGRATION

Keep the tools you know.

Connect Puppeteer, Playwright, or your own CDP client. Your existing automation has a place to run.

ORCHESTRATION

One place to orchestrate.

Manage sessions, profiles, and browser instances together. Let the orchestrator handle allocation.

02 / FROM INTENT TO ACTION

Three steps.
A browser that gets to work.

See the code
01

Create a session

Send a request to the API. Set your profile, proxy, or timezone for the task ahead.

POST /api/v1/sessions
02

Connect over CDP

Use the returned WebSocket URL with Puppeteer, Playwright, or your own agent.

session.cdp_url
03

Let your agent work

Navigate, click, and extract. Watch in the console and close the session when you’re done.

DELETE /api/v1/sessions/:id
03 / DEVELOPER FIRST

Your stack.
Your code.
Our browsers.

No new automation language to learn. Create a session, connect your client, and keep building with the tools you already use.

  • Standard CDP WebSocket connection
  • REST API for session lifecycle
  • Works with existing API keys
Read the API reference

Already have access? Use your API key.
New keys are provided by your 0Browser operator.

// npm install puppeteer-core (Node.js 22+)import puppeteer from "puppeteer-core"; const api = "https://api.0browser.io/api/v1/sessions";const headers = {  Authorization: `Bearer ${process.env.ZERO_API_KEY}`,  "Content-Type": "application/json",};const response = await fetch(api, {  method: "POST", headers, body: JSON.stringify({ os: "linux" }),});if (!response.ok) throw new Error(await response.text());const session = await response.json();let browser;try {  browser = await puppeteer.connect({    browserWSEndpoint: session.cdp_url,  });  const page = await browser.newPage();  await page.goto("https://example.com");  console.log(await page.title());} finally {  browser?.disconnect();  const closed = await fetch(`${api}/${session.id}`, {    method: "DELETE", headers,  });  if (!closed.ok) throw new Error(await closed.text());}
A FEW MORE DETAILS

Good questions.
Clear answers.

How do I get started?

Open the console and sign in with your existing 0Browser API key. You can create a browser session there, or use the API examples above. New users need a key from their 0Browser operator; self-service signup is not available yet.

Can I use Puppeteer or Playwright?

Yes. Each session returns a CDP WebSocket URL. Connect using Puppeteer’s browserWSEndpoint or Playwright’s chromium.connectOverCDP, then use your usual browser automation methods.

What persists between sessions?

Persistent profiles reuse browser data such as cookies and local storage across runs. Select an existing profile when creating a session to continue with its saved state. Sessions without a profile are ephemeral.

Can I watch or interact with a session?

Yes. Open the session in the console to use the live viewer. You can inspect the page and interact with the browser while your automation runs.

Can I configure proxies and browser settings?

You can provide a proxy and timezone when creating a session, and manage browser fingerprint settings through profiles. The console brings these settings together in one place.

THE WEB IS YOUR NEXT WORKSPACE

Give your agent
a window to the web.

Your next browser session starts here.

Open Console Sign in with your existing API key.