> ## Documentation Index
> Fetch the complete documentation index at: https://docs.openbunny.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Quickstart

> Get OpenBunny running locally in under 5 minutes.

This guide walks you through setting up OpenBunny in fully local mode with SQLite. You'll have a working task extraction system connected to at least one messaging channel.

## Prerequisites

* Node.js 22 or later
* An [OpenRouter](https://openrouter.ai) API key
* Credentials for at least one messaging channel (Slack, WhatsApp, Telegram, or Gmail)

## Setup

<Steps>
  ### Clone the repository

  ```bash theme={null}
  git clone https://github.com/openbunny/openbunny.git
  cd openbunny
  ```

  ### Install dependencies

  ```bash theme={null}
  cd service
  npm install
  ```

  ### Configure environment

  Copy the example environment file and add your API key:

  ```bash theme={null}
  cp .env.example .env
  ```

  ```bash .env theme={null}
  BACKEND=local
  SQLITE_PATH=./data/openbunny.db
  OPENROUTER_API_KEY=sk-or-v1-your-key-here
  OPENROUTER_MODEL=anthropic/claude-sonnet-4
  ```

  ### Connect a channel

  Run the interactive setup to configure your first channel:

  ```bash theme={null}
  npm run setup
  ```

  This walks you through connecting Slack, WhatsApp, Telegram, or Gmail. See the [Channels](/service/channels/slack) docs for detailed setup instructions per channel.

  ### Start the service

  ```bash theme={null}
  npm run dev
  ```

  The service starts on `http://localhost:3100`. It begins listening to your configured channels immediately.

  ### Start the client (optional)

  In a new terminal:

  ```bash theme={null}
  cd ../client
  npm install
  npm run dev
  ```

  Open `http://localhost:3000` to see the Kanban board UI.
</Steps>

## Verify it's working

Check the service health endpoint:

```bash theme={null}
curl http://localhost:3100/health
```

Send a message in your connected channel that contains an actionable task. Within a few minutes (after the debounce window), OpenBunny will process the conversation and create a task.

View your tasks:

```bash theme={null}
curl http://localhost:3100/tasks | jq
```

## Next steps

<CardGroup cols={2}>
  <Card title="Architecture" icon="sitemap" href="/architecture">
    Understand how the components fit together
  </Card>

  <Card title="Configuration" icon="gear" href="/service/configuration">
    Fine-tune the service behavior
  </Card>

  <Card title="Deploy to cloud" icon="cloud" href="/deployment/hybrid">
    Set up multi-tenant cloud deployment
  </Card>

  <Card title="LLM agent" icon="robot" href="/service/agent">
    Learn how the agent processes conversations
  </Card>
</CardGroup>