Browserbase integrations

Browserbase runs cloud browsers for agents, with session recordings and a live view for debugging. The same API key covers Search and Fetch, which find and read pages without starting a browser session. This page covers using Browserbase with Deep Agents: a Stagehand-driven cloud browser, plus Search and Fetch as agent tools.

Installation and setup

Get an API key from Browserbase and set it as an environment variable:

export BROWSERBASE_API_KEY="your-browserbase-api-key"

Install the Browserbase Python SDK:

pip install "browserbase>=1.21.0"

uv add "browserbase>=1.21.0"

Deep Agents

A deep agent can use Browserbase in two ways: a cloud browser that it drives through Stagehand tools, and Search and Fetch tools for pages that do not need a browser. For when to give an agent a browser, see Browser use.

Browser tools with Stagehand

The Stagehand integration adds run, snapshot, and screenshot through native Python middleware. Pass a configured Stagehand client to StagehandMiddleware, then include it in your agent’s middleware list. No MCP server or custom browser system prompt is needed. Your application owns the browser and can keep it open across runs to reuse tabs and logins. For installation, a complete agent, and cleanup, see Use with a deep agent. For session isolation, see Choose a browser scope.

Search and fetch tools

Expose Search and Fetch as agent tools. Those APIs find and read pages without starting a browser session. Fetch does not execute JavaScript. Install the agent packages and set the model-provider credential before running this example:

pip install "browserbase>=1.21.0" deepagents langchain-openai

uv add "browserbase>=1.21.0" deepagents langchain-openai

export OPENAI_API_KEY="your-openai-api-key"

The example uses OpenAI’s GPT-5.5 snapshot. Omit provider on search.web to use your project’s default search provider:

import os

from browserbase import Browserbase
from deepagents import create_deep_agent
from langchain.tools import tool

bb = Browserbase(api_key=os.environ["BROWSERBASE_API_KEY"])


@tool
def web_search(query: str) -> str:
    """Search the web and return result titles and URLs."""
    response = bb.search.web(query=query, num_results=5)
    return "\n".join(f"{result.title}: {result.url}" for result in response.results)


@tool
def fetch_page(url: str) -> str:
    """Fetch a page as markdown without opening a browser."""
    return bb.fetch_api.create(url=url, format="markdown").content


agent = create_deep_agent(
    model="openai:gpt-5.5-2026-04-23",
    tools=[web_search, fetch_page],
    system_prompt="Search to find pages, then fetch the pages you need to read.",
)

For a complete example that combines Search, Fetch, and Stagehand, see Browserbase’s Deep Agents guide.

Connect these docs to your agent of choice via MCP for real-time answers.

Edit this page on GitHub or file an issue.

Popular Providers

Integrations by component

Installation and setup

Deep Agents

Browser tools with Stagehand

Search and fetch tools

​Installation and setup

​Deep Agents

​Browser tools with Stagehand

​Search and fetch tools

Installation and setup

Deep Agents

Browser tools with Stagehand

Search and fetch tools