Does it cost more than running Chromium locally?

Yes — Browserbase charges per browser-minute. The trade is reliability, parallelism, and IP diversity. For a few sessions a day, local Puppeteer is cheaper. For dozens of concurrent agents, hosted is usually less hassle.

Can sessions persist across runs?

Yes. The MCP exposes session IDs that survive between tool calls and across MCP restarts. Useful for "log in once, run multiple workflows" patterns where you do not want to re-auth every time.

How does it compare to Stagehand?

Stagehand is a higher-level layer that adds AI primitives (`act`, `observe`, `extract`) on top of Browserbase. If you want natural-language browser commands, install the Stagehand MCP; if you want raw Playwright-style control, the Browserbase MCP is enough.

Browserbase

Name: Top MCPs — curated MCP directory
Creator: Top MCPs
License: https://creativecommons.org/licenses/by/4.0/

Official

Hosted, isolated Chromium runtime for AI agents that need a fresh browser per task.

Score 89(?)BrowserbaseMIT 3.4kVerified May 26, 2026Top MCPs for Browser Automation →

Install View on GitHub

Quick answer

What it does

Provisions cloud Chromium sessions on demand. Exposes navigate, click, type, screenshot, and cookie-persistence tools, with optional Stagehand AI primitives on top.

Best for

Hosted browser sessions
Concurrent automation at scale
Anti-bot resilience
Persistent sessions across runs

Not for

Offline workflows
Hobbyist budgets

Setup recipe

Pick your client, then follow the three steps.

Install

claude_desktop_config.jsonjson

{
  "mcpServers": {
    "browserbase": {
      "command": "npx",
      "args": [
        "-y",
        "@browserbasehq/mcp"
      ],
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

Paste under mcpServers. Fully quit and reopen Claude after editing.

CLI or .mcp.jsonshell

# export BROWSERBASE_API_KEY=YOUR_API_KEY
# export BROWSERBASE_PROJECT_ID=YOUR_PROJECT_ID
claude mcp add browserbase -- npx -y @browserbasehq/mcp

Run from your repo. Commit .mcp.json to share with your team.

.cursor/mcp.jsonjson

{
  "mcpServers": {
    "browserbase": {
      "command": "npx",
      "args": [
        "-y",
        "@browserbasehq/mcp"
      ],
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

Global path: ~/.cursor/mcp.json. Reload window after editing.

.vscode/mcp.jsonjsonc

{
  "servers": {
    "browserbase": {
      "command": "npx",
      "args": [
        "-y",
        "@browserbasehq/mcp"
      ],
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

VS Code uses the "servers" key (not "mcpServers").

~/.codeium/windsurf/mcp_config.jsonjson

{
  "mcpServers": {
    "browserbase": {
      "command": "npx",
      "args": [
        "-y",
        "@browserbasehq/mcp"
      ],
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

Open via Cascade → hammer icon → Configure.

cline_mcp_settings.jsonjson

{
  "mcpServers": {
    "browserbase": {
      "command": "npx",
      "args": [
        "-y",
        "@browserbasehq/mcp"
      ],
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

Open via the Cline sidebar → MCP Servers → Edit.

~/.continue/config.jsonjson

{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "transport": {
          "type": "stdio",
          "command": "npx",
          "args": [
            "-y",
            "@browserbasehq/mcp"
          ],
          "env": {
            "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
            "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
          }
        }
      }
    ]
  }
}

Continue uses modelContextProtocolServers with a transport block.

~/.codex/config.tomlshell

# ~/.codex/config.toml
[mcp_servers.browserbase]
command = "npx"
args = [
  "-y",
  "@browserbasehq/mcp",
]
env = { BROWSERBASE_API_KEY = "${BROWSERBASE_API_KEY}", BROWSERBASE_PROJECT_ID = "${BROWSERBASE_PROJECT_ID}" }

Codex uses TOML. Each server is a [mcp_servers.<name>] subtable.

~/.config/zed/settings.jsonjsonc

{
  "context_servers": {
    "browserbase": {
      "command": {
        "path": "npx",
        "args": [
          "-y",
          "@browserbasehq/mcp"
        ]
      },
      "env": {
        "BROWSERBASE_API_KEY": "${BROWSERBASE_API_KEY}",
        "BROWSERBASE_PROJECT_ID": "${BROWSERBASE_PROJECT_ID}"
      }
    }
  }
}

Zed calls them "context_servers". Settings live-reload on save.

ChatGPT → Apps directorynone

Browserbase doesn't ship a hosted HTTPS endpoint today. ChatGPT supports remote MCP servers only — to use this server in ChatGPT you'll need to deploy it to a public HTTPS URL first (e.g. via Cloudflare Workers or Vercel) or wait for an official remote build.

2
Set required secrets
Set BROWSERBASE_API_KEY, BROWSERBASE_PROJECT_ID in your shell environment before launching your MCP client.
3
Try a minimum working prompt
Minimum working prompt pending verification. Try any prompt from the MCP’s README once installed.

Tools & permissions

Tools list pending verification. The server exposes tools over MCP; we haven’t yet parsed its capability manifest into this page. Check the GitHub repo for the authoritative list.

Security & scope

Access scope: Exec
Sandbox: Each session runs in an isolated cloud container under your Browserbase account. The browser has internet access but no access to your local filesystem.
Gotchas: Captured screenshots and downloads live on Browserbase's side until you fetch them.
API key grants project-wide access; rotate if leaked.
Sessions left open accrue per-minute charges — always close.

Agent prompt pack

— copy into Claude, Cursor, or ChatGPT.

Paste into Claude, Cursor, or ChatGPT. Edit the [brackets] before sending.

Recommend the best MCP servers for [task: e.g. browser automation work] in [client: Claude].

Constraints:
- Prefer tools that are [official | open-source | read-only] — pick what matters for my use case.
- Exclude MCPs that require [e.g. a paid plan, OAuth-only flows, remote-only transport].
- Return at most 3 picks, ranked.

For each pick include:
1. One-sentence rationale.
2. The ready-to-paste install snippet for my client.
3. Any required secrets I need to create before installing.

Cross-check the top-mcps.com listing: https://top-mcps.com/top-mcps-for-browser-automation

Compare Browserbase against a real alternative. Swap the second MCP in [brackets] if you want a different match.

Compare Browserbase MCP vs [Playwright MCP] for the following job: [describe the job, e.g. "let an agent create GitHub issues on bug triage"].

Judge them on:
- Setup time and complexity (what a new user hits first).
- Auth model (none / API key / OAuth 2.1) and credential risk.
- Transport (stdio / Streamable HTTP / SSE) and where the server runs.
- Required secrets and the blast radius if they leak.
- Operational risk in an unattended agent loop.
- Which one is "good enough" for a weekend prototype vs. production.

End with one sentence: which should I pick for my scenario, which is: [my scenario].

References:
- https://top-mcps.com/mcp/browserbase
- top-mcps.com listing for Playwright

Asks the agent to install and verify. Works inside Claude Code, Cursor Agent, Codex CLI.

Install the Browserbase MCP server for my [client: Claude] at the default config path for that client.

Use the exact install snippet published at https://top-mcps.com/mcp/browserbase (fetch https://top-mcps.com/mcp/browserbase.json for the canonical server.json if you can read URLs).

Before finishing:
1. Create the required secrets (BROWSERBASE_API_KEY, BROWSERBASE_PROJECT_ID) and put them in the appropriate env block — do not hard-code them.
2. Restart or reload the client so it picks up the new server.
3. Verify the server is connected (green / running state) and at least one tool is listed.
4. If anything fails, read the client's MCP logs and report the exact error — do not silently retry.

Confirm when done and list the tools the server now exposes.

Frequently asked questions

What changed

— 2 updates tracked.

May 26, 2026
Refreshed install snippets and fact sheet; verified for 2026.
Nov 30, 2024
Initial directory listing.

More Browser Automation MCPs

Other tools in the same category worth evaluating.

Puppeteer

Full browser automation: navigate, click, screenshot, and scrape.

browserautomationscrapingpuppeteer

5 minMedium

Playwright

Official

Official Microsoft browser automation across Chromium, Firefox, and WebKit.

browserautomationplaywrighttesting

5 minMedium

Stagehand

Official

AI-native browser automation: act, observe, and extract in plain English.

browserstagehandbrowserbaseai-automation

5 minMedium

Compared with Browserbase

Side-by-side breakdowns for the choices people most often weigh against this MCP.

BrowserbasevsPlaywright BrowserbasevsPuppeteer BrowserbasevsStagehand BrowserbasevsSkyvern

Exploring Top MCPs for Browser Automation? See all Browser Automation MCPs →

Quick answer

What it does

Best for

Not for

Setup recipe

Install

Set required secrets

Try a minimum working prompt

Tools & permissions

Security & scope

Agent prompt pack

Full details

Frequently asked questions

Does it cost more than running Chromium locally?

Can sessions persist across runs?

How does it compare to Stagehand?

What changed

More Browser Automation MCPs

Puppeteer

Playwright

Stagehand

Compared with Browserbase