# LLM Arena

Compare AI models by executing identical tasks simultaneously.

## The Idea

Which AI model performs best for browser automation tasks? The LLM Arena template facilitates **side-by-side comparison** of AI models by executing identical tasks across each simultaneously and ranking results by performance.

## Requirements

You'll need **four API keys** for full functionality:

- **Browser Use** (Required)
- **Google Gemini** (Optional but recommended)
- **OpenAI** (Optional but recommended)
- **Anthropic** (Optional but recommended)

Get your Browser Use API Key [here](https://cloud.browser-use.com/new-api-key)

Get your Google API Key [here](https://aistudio.google.com/app/apikey)

Get your OpenAI API Key [here](https://platform.openai.com/api-keys)

Get your Anthropic API Key [here](https://console.anthropic.com/settings/keys)

**Note:** You can run with just Browser Use, but for true comparison configure all four providers.

## Installation

```bash
uvx browser-use init --template llm-arena
```

## How It Works

1. Input a task description via CLI  
2. Launches parallel executions across all configured LLMs  
3. Each model operates independently with timing tracked  
4. Results display with performance rankings from fastest to slowest

To ensure fair comparisons, this template uses the `@sandbox()` decorator to run tasks in Browser Use Sandboxes.

## Supported Models

By default, the template compares:

- **Browser Use**  
- **Google Gemini**  
- **OpenAI GPT**  
- **Anthropic Claude**