Skip to content
  1. Home
  2. AI tools
  3. Stagehand

Stagehand

SDK for browser agents: act, extract and observe pages in plain language

What it is

A TypeScript and Python SDK that mixes ordinary browser automation code with AI steps. You can ask it to act on a page, extract structured data against a schema or observe elements to get real selectors, and it repairs steps when a site changes. For developers who want controllable browser agents rather than fully autonomous ones.

Good for

  • Extract structured data from a logged-in page
  • Write resilient end-to-end browser steps
  • Keep credentials out of model prompts

Install or run

pnpm add @browserbasehq/stagehand 'zod@~4.4.3'

Command as given in the project README. Read the repository before you run anything on your machine.

What you need

Node.js or Python, an LLM API key; local Chrome or a Browserbase account

Worth knowingMaintained by Browserbase and its docs lean toward that paid cloud browser, though local browsers work.

Sources

The summary is ours, written from the README. Spotted an error? See how we work.

More for coding

See the topic →
  • opencode

    anomalyco/opencode

    Open-source AI coding agent with a terminal UI and desktop app

    ToolGitHub stars: 227k
  • Claude Code

    anthropics/claude-code

    Anthropic's terminal coding agent; this repo hosts issues and plugins

    ToolGitHub stars: 156k
  • Spec Kit

    github/spec-kit

    GitHub's toolkit for spec-driven development with coding agents

    ToolGitHub stars: 150k
  • Codex CLI

    openai/codex

    OpenAI's terminal coding agent that runs locally on your machine

    ToolGitHub stars: 135k