> For the complete documentation index, see [llms.txt](https://docs.varbase.vardot.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.varbase.vardot.com/developers/understanding-varbase/varbase-ai-recipes/varbase-ai-chatbot-base.md).

# Varbase AI Chatbot Base

The **Varbase AI Chatbot Base** recipe adds a chatbot for site visitors. It answers only from the site's own content, links the pages it used, and says so when the site has no answer.

## Recipe Type

Varbase AI

## Drupal.org Project

<https://www.drupal.org/project/varbase_ai_chatbot_base>

{% hint style="warning" %}
**Alpha release.** This recipe is not covered by the Drupal security advisory policy. Try it on a development site first.
{% endhint %}

## What It Adds

* A **Varbase AI chatbot** agent that searches the `content_vector` index before every answer.
* An **AI Assistant** for the chatbot block, open to anonymous and authenticated visitors.
* The `access deepchat api` permission for anonymous and authenticated users.

## Before You Apply

Apply in this order: [**Varbase AI Base**](/developers/understanding-varbase/varbase-ai-recipes/varbase-ai-base.md), one vector database recipe, then [**Varbase AI Search Base**](/developers/understanding-varbase/varbase-ai-recipes/varbase-ai-search-base.md), and index the content.

## Installation

Add the recipe with Composer, then apply it with Drush:

```bash
ddev composer require drupal/varbase_ai_chatbot_base:~1
ddev drush cr
ddev drush recipe ../recipes/varbase_ai_chatbot_base
ddev drush cr
```

## Usage

The chatbot is the **AI DeepChat Chatbot** block, with its assistant set to **Varbase AI chatbot**. Keep streaming off on the block.

{% hint style="warning" %}
On a Drupal Canvas theme the chatbot block cannot be placed yet: Canvas does not accept the block as a component. On a theme that renders block regions, place it from **Block layout** as usual.
{% endhint %}

Varbase's consent manager asks visitors before the chatbot loads, because their questions are sent to the AI provider.

On a theme that renders block regions, use these block settings: placement **Bottom right**, streaming **off** (with streaming on, the retrieved context was dropped mid-answer at times), and a first message saying the chatbot answers only from the site's pages.

## Where Data Goes

* The content stays in Drupal and in the vector database the site owner chooses.
* Indexing sends each content chunk to the embeddings provider once, and again when the content changes.
* Each answer sends the visitor's question and the retrieved passages to the chat model provider. Choose a provider whose data terms fit the site, or a self-hosted model.