Browser Use

Browser Use ships as three layered products: the open-source browser-use Python library (an agent loop that converts a goal into browser actions, for self-hosted use with your own LLM keys), a managed cloud browser service (remote Chromium sessions with stealth/proxy/profile features, drivable via the standard Playwright/Puppeteer/Selenium APIs or the vendor SDK), and a hosted Agent API (v4) that accepts a natural-language task and returns a structured result without the caller writing any browser-automation code itself.

• Hosted agents: natural-language task in, structured JSON-schema output back, with multi-turn sessions and persistent context
• Workspace file management and deterministic script replay of agent runs
• Human-in-the-loop intervention for authentication steps (2FA, login walls)
• Managed cloud browser infrastructure: remote Chromium controllable via Python/TypeScript SDK or standard Playwright/Puppeteer/Selenium
• Anti-detection stealth browser builds and global residential proxy coverage (195+ countries)
• Session persistence, live preview, and session recordings
• Open-source Python agent loop (browser-use library) with multi-LLM provider support and customizable tools/prompts/schemas
• Webhooks and observability for agent runs

Developers building AI agents or automation pipelines that need browser access - ranging from teams that just want a hosted API to run a task, to teams that want raw cloud browser infrastructure under their own automation code, to teams that want to self-host the open-source agent library entirely.

Pricing is entirely usage-metered (browser-hours, proxy GB, and token costs for the hosted agent model) with no flat-rate plan below Enterprise, so a team with unpredictable or bursty automation volume cannot get a fixed monthly bill without negotiating an Enterprise contract. Managed residential proxy bandwidth is billed separately at $5/GB (or $4/GB on Scaleup) on top of browser-hour and token costs, which stacks up for proxy-heavy scraping workloads compared to flat-fee scraping APIs. No MCP server documentation was found, so teams specifically wanting an out-of-the-box MCP integration for AI assistants will need to build one against the REST API themselves.

Setup assessment: 5m

The cloud Agent API quickstart is: create an account, generate an API key from the dashboard, export BROWSER_USE_API_KEY, and send a POST /runs request - no OAuth flow. The open-source library and cloud browser infrastructure paths take longer to configure (LLM provider keys, proxy/profile setup) but are not required for the basic hosted-agent flow.

Prerequisites

• Browser Use Cloud account and API key
• For the open-source library: an LLM provider API key (OpenAI, Anthropic, etc.)

Authentication

API key sent via the X-Browser-Use-API-Key header

Base URL

https://api.browser-use.com/api/v4

Machine-readable API

OpenAPI specification linked

• POST /runs - create and execute a hosted agent task
• GET /runs/{id} - retrieve agent run status, events, and attachments
• POST /browsers - provision a cloud browser session
• PATCH /browsers/{id} - control browser session lifecycle
• /sessions - manage conversation sessions with message queuing
• /workspaces - file upload and workspace management
• /profiles - create and manage browser profiles for preserving login state

• No MCP server documentation was found on the docs site
• No official YouTube channel could be confirmed (a channel handle exists but page content did not confirm ownership or show videos)
• No LinkedIn or X/Twitter links were found in the site footer or docs at check time
• No public changelog page was found
• GitHub star counts beyond the figures visible on the org's repo list at check time