AI agents / Anti-detect browser for AI agents

An anti-detect browser for AI agents that limits what each agent can do.

You decide exactly what each agent may do: its own key, its own profiles, only the sites you allow, sign-ins it can use on your computer but never read, and a person for the login. Each agent task runs in its own isolated profile, with one consistent device fingerprint, its own proxy and its own cookies, so one task’s sign-ins are never there for the next, and your agent drives it through the bundled MCP server. Underneath, every profile’s fingerprint is set in the browser engine itself, not by a script on top, and you can test it yourself (engine-level vs injected scripts).

Set it up Download Browz

MCP client config

{
  "mcpServers": {
    "browz": {
      "command": "/Applications/Browz.app/Contents/Resources/browz-mcp",
      "args": []
    }
  }
}

Anti-detect browser for AI agents

What your agent gets.

You decide what each agent may do

Each agent gets its own key, limited to the profiles and sites you choose, with call limits, a log on your computer, revoke and “stop all agents”. On your computer it can sign in with a password and code you stored without ever seeing them, and it hands logins, 2FA and CAPTCHAs to a person.

Set an agent’s limits 02

One consistent device

The device a site reads is set inside the browser engine, not with injected scripts, and stays the same from visit to visit. The TLS fingerprint (JA4) matches stock Google Chrome’s, and driving a profile over MCP doesn’t change its fingerprint.

See the protections 03

A proxy that matches

Each profile has its own checked proxy exit, with the timezone, language and location following it, so an agent in Sydney looks like it is in Sydney.

About proxies 04

Measured, then published

In our published benchmark the same AI agent finished 46 of 50 public web tasks through Browz and 38 of 50 through a plain browser MCP server; on sites that challenge automated browsers it was 14 of 18 against 9 of 18.

See the benchmark

Setup

Working in minutes.

These are the same snippets as the setup docs and the Connect an AI agent screen in the app.

Install Browz and open it
The MCP server ships inside the app and drives its profiles. Download Browz for Mac, Windows or Linux; it is free for 5 profiles.
Add Browz to your agent
Paste the config into any MCP client that runs a local server, such as Claude Desktop, Claude Code, Cursor, VS Code or Gemini CLI. Hosted agents add the remote server instead, in beta.
Give it a task
For example: “Using Browz: create a new profile called agent-check, start it, open https://example.com, tell me the page title, take a screenshot, then stop the session and delete the profile.” It should answer “Example Domain”.

Install Browz, open it, and your agent can start: no account needed for local use. Keep the app open while your agent works; it can stay on its sign-in screen.

MCP client config Add this entry to your client’s MCP config file (for Claude Desktop, claude_desktop_config.json; for Cursor, mcp.json), then restart the client.

MCP client config

{
  "mcpServers": {
    "browz": {
      "command": "/Applications/Browz.app/Contents/Resources/browz-mcp",
      "args": []
    }
  }
}

Hosted agents (beta) Add this URL as a remote MCP server and sign in with your Browz account.

Hosted agents (beta)https://mcp.gobrowz.com/mcp

The snippets use the macOS path to the server inside Browz, /Applications/Browz.app/Contents/Resources/browz-mcp; on Windows and Linux the paths are in the setup docs. Give each agent its own key, limited to the profiles and sites it needs, and let it sign in without seeing the password. The remote MCP server for hosted agents is in Beta: free for every account, Free included, within the beta allowance, and it needs the Browz app open with Remote control on, on a computer with a hardware security chip. See what it needs and its allowance. Use agents only where each site’s terms allow it: our acceptable use policy applies to agents the same as to people.

FAQ

About Anti-detect browser for AI agents.

What is an anti-detect browser for AI agents?

A browser that gives each agent task a separate, consistent identity, its own device fingerprint, proxy and cookies, so tasks never share accounts or sessions. Use it only where a site’s terms allow automated use. Browz adds an MCP server so the agent can create and drive those profiles itself.

Why not let the agent use my own browser?

An agent in your own browser can reach every account you are signed in to, and a page that tricks it can act on all of them. In Browz each task has a profile of its own, and your everyday browser is never part of it.

How does Browz compare with other browsers for AI agents?

We scored ten browsers that agents use on thirteen published checkpoints, from each vendor’s public pages, with every source linked: the best anti-detect browsers for AI agents. On how the fingerprint itself is made, see engine-level vs injected scripts.

What happens when a site shows the agent a CAPTCHA?

You complete it yourself in the profile’s window and the agent carries on in the same session. CAPTCHAs always come to a person: Browz never solves or bypasses them. The agent can also pause and hand the window to you, or sign in with a password and code you stored without seeing them.

More for agent builders

Your agent. Its own browser.

Download Browz See pricing

For Mac, Windows and Linux. Free for 5 profiles.

An anti-detect browser for AI agents that limits what each agent can do.

What your agent gets.

You decide what each agent may do

One consistent device

A proxy that matches

Measured, then published

Working in minutes.

Install Browz and open it

Add Browz to your agent

Give it a task

About Anti-detect browser for AI agents.

Other guides.

Browz for AI agents

MCP browser

Browz for Claude

Browz for Cursor

Your agent. Its own browser.