BROWSER USE

- Browser Use Agents: give Browser Use a task and receive completed work. API V4 is current for new integrations.
- Browser Infrastructure: connect your agent or automation to managed browsers through SDK, REST, or CDP. Starts at $0.02/browser-hour.
- Developer tools: Open Source, Browser Harness, SDK, and MCP support the two products above.

[Developer Index](https://browser-use.com/index.md)
[Product Map](https://browser-use.com/llms.txt)
[Full Product Context](https://browser-use.com/llms-full.txt)
[Pricing](https://browser-use.com/pricing.md)
[Cloud Docs](https://docs.browser-use.com/cloud/quickstart)
[Open Source Docs](https://docs.browser-use.com/open-source/introduction)

---

# Advanced Configuration

> Full control over Browser Use agent behavior.

---

### The Idea

When getting started with Browser Use, the default template provides a quick setup. But what if you need **full control** over agent and browser behavior? 

The advanced template highlights recommended advanced configuration options for both the Browser and Agent, making it a **comprehensive reference guide** for users.

### Requirements

You'll need one API Key:
- **Browser Use**



### Installation


uvx browser-use init --template advanced


### What's Included

**Browser Configuration:**
- `headless` - Run browser with or without UI (auto-detects based on display)
- `cdp_url` - Connect to existing browser instance via Chrome DevTools Protocol
- `allowed_domains` / `prohibited_domains` - Restrict navigation to specific domains using pattern matching
- `args` - Pass custom command-line arguments to Chrome
- `keep_alive` - Keep browser running after agent completes
- `user_data_dir` - Specify browser profile directory (None for incognito mode)
- `proxy` - Configure proxy settings for network requests
- `wait_between_actions` - Control timing between agent actions
- `highlight_elements` - Highlight interactive elements for AI vision

**Agent Configuration:**
- `tools` - Registry of custom tools the agent can call
- `output_model_schema` - Pydantic model for structured output validation
- `use_vision` - Vision mode ('auto', True, False) for screenshot handling
- `max_actions_per_step` - Maximum actions per step (e.g., form fields)
- `max_failures` - Maximum retries for steps with errors
- `save_conversation_path` - Path to save complete conversation history
- `generate_gif` - Generate GIF of agent actions (True or path string)
- `step_timeout` - Timeout in seconds for each step
- `llm_timeout` - Timeout in seconds for LLM calls
- `calculate_cost` - Calculate and track API costs

### Customization

All parameters are set to their default values and include inline comments. Simply uncomment and modify any option to customize behavior for your specific use case. For detailed documentation on each parameter, visit the [ Browser Use docs](https://docs.browser-use.com/introduction).