index.md

# Hyperbrowser

## AI Agents Gateway to the Internet

Automate websites, run code in isolated environments, and extract web data with APIs and SDKs.

[Start for free](https://app.hyperbrowser.ai/signup) · [Book a demo](https://calendly.com/shri-hyperbrowser/demo)

## Get started

- [Documentation](https://www.hyperbrowser.ai/docs)
- [Documentation index for agents](https://www.hyperbrowser.ai/docs/llms.txt)
- [Browser quickstart](https://www.hyperbrowser.ai/docs/quickstart)

### Install the CLI

```sh
curl -fsSL https://www.hyperbrowser.ai/cli/install.sh | sh
```

CLI installation: https://www.hyperbrowser.ai/docs/sandboxes/install

### Agent setup prompt

Set up Hyperbrowser in this project so my agent can use cloud browsers and sandboxes.

Read the official documentation first:
- Documentation index: https://www.hyperbrowser.ai/docs/llms.txt
- CLI installation: https://www.hyperbrowser.ai/docs/sandboxes/install
- CLI commands: https://www.hyperbrowser.ai/docs/sandboxes/cli
- Browser quickstart: https://www.hyperbrowser.ai/docs/quickstart

1. Inspect the project and follow its existing language, package manager, and configuration conventions.
2. Install the hx CLI if it is missing. On macOS or Linux, use:
   curl -fsSL https://www.hyperbrowser.ai/cli/install.sh | sh
   On Windows, follow the official Windows installation instructions instead. Verify with hx version and hx --help.
3. Run hx auth login and let me complete the interactive sign-in. Reuse an existing authenticated profile when available.
4. For browser automation, install the official SDK for this project's language and add a minimal example that opens https://example.com, reads the page title, and closes the session in a finally block. Load HYPERBROWSER_API_KEY from the environment; keep credentials out of source control and logs.
5. Explain how to run the example and how to use hx for sandbox workflows. Summarize what you changed and any remaining setup steps.

## The platform

Browsers, compute, and web data.

### Browsers

Control Chrome through an API

Connect Playwright or Puppeteer to a cloud-hosted Chrome session. Save logins between runs and inspect sessions live.

- Stealth & residential proxies
- Persistent profiles
- Live view & recordings

[Explore Browsers](https://www.hyperbrowser.ai/browser-sessions) · [Documentation](https://www.hyperbrowser.ai/docs/sessions/create)

### Sandboxes

Run code in isolated environments

Install dependencies, execute commands, and expose local services. Save the environment as a snapshot to reuse in later runs.

- Isolated virtual machines
- Exposed ports
- Memory snapshots

[Explore Sandboxes](https://www.hyperbrowser.ai/sandboxes) · [Documentation](https://www.hyperbrowser.ai/docs/sandboxes/introduction)

### Web API

Extract data from websites

Fetch a page, crawl a site, or search the web. Return Markdown, HTML, screenshots, or JSON that matches your schema.

- Fetch, crawl & search
- Schema-based extraction
- Screenshots

[Explore Web API](https://www.hyperbrowser.ai/web-apis) · [Documentation](https://www.hyperbrowser.ai/docs/web/overview)

### Agents

Run browser tasks from a prompt

Run HyperAgent, Browser Use, or computer-use models in a managed browser. Send a task, inspect the steps, and retrieve the result.

- Natural-language tasks
- Managed execution
- Structured results

[Explore Agents](https://www.hyperbrowser.ai/ai-agents) · [Documentation](https://www.hyperbrowser.ai/docs/agents/overview)

## In production

Session state, debugging, and isolation.

### Saved state

Reuse browser profiles with saved cookies and logins. Restore sandbox environments from memory snapshots.

[Learn more](https://www.hyperbrowser.ai/docs/sessions/profiles)

### Session recordings

Watch browser sessions live and replay recordings to debug a run.

[Learn more](https://www.hyperbrowser.ai/docs/sessions/recordings)

### Isolated execution

Run commands in isolated virtual machines, separate from your application.

[Learn more](https://www.hyperbrowser.ai/sandboxes)

### SDKs and integrations

Use Playwright, Puppeteer, the Node.js or Python SDKs, the CLI, or MCP.

[Learn more](https://www.hyperbrowser.ai/docs/quickstart)

### Security and access controls

Read about encryption, access controls, and independent security assessments.

[Security overview](https://www.hyperbrowser.ai/security)

## Pricing

Start with 5,000 free credits.

- **Free**: $0. 5,000 credits.
- **Startup**: $30/month. 30,000 credits.
- **Scale**: $100/month. 100,000 credits.
- **Enterprise**: Custom. Contact sales.

[All plans and usage rates](https://www.hyperbrowser.ai/pricing)

## From the team

Blog

- [Introducing WebMCP Support](https://www.hyperbrowser.ai/blog/webmcp-on-hyperbrowser)
- [Browser sessions with x402: Let agents buy their own browser](https://www.hyperbrowser.ai/blog/x402-browser-sessions)
- [Introducing Hyperbrowser MCP Server: AI Meets the Web](https://www.hyperbrowser.ai/blog/introducing-hyperbrowser-mcp-server)

[More from the blog](https://www.hyperbrowser.ai/blog)

## Start with a free account.

Includes 5,000 credits for browsers, sandboxes, and web data.

[Create an account](https://app.hyperbrowser.ai/signup) · [Read the docs](https://www.hyperbrowser.ai/docs/quickstart)

---

[Human version](https://www.hyperbrowser.ai/) · [Raw Markdown](https://www.hyperbrowser.ai/index.md)