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

# Models

> Configure and manage AI models for chat, embedding and voice

<Tip>
  Home > "Admin Menu" > "Models"
</Tip>

The model management provides a central interface for configuring all AI models in VARIOS AI. Models are divided into three categories:

| Category             | Description                                                               |
| -------------------- | ------------------------------------------------------------------------- |
| **Chat Models**      | Models for text generation and conversations (e.g. GPT, Claude, DeepSeek) |
| **Embedding Models** | Models for text vectorization and semantic search                         |
| **Voice Models**     | Models for speech-to-text (transcription) and text-to-speech (speech)     |

## Overview

Use the **tabs** at the top to switch between the three model categories. Each category displays its own list of configured models.

* **Search field**: Filter models by name.
* **New Model**: Click the button in the top right to add a new model.

### Actions per Model

| Action             | Description                           |
| ------------------ | ------------------------------------- |
| **Edit** (Pencil)  | Open and edit the model configuration |
| **Delete** (Trash) | Permanently remove the model          |

## Detailed Configuration

<CardGroup cols={3}>
  <Card title="Chat Models" icon="message" href="./models/chat">
    Text generation, conversations, Code Interpreter – all fields and options in
    detail.
  </Card>

  <Card title="Embedding Models" icon="vector-square" href="./models/embedding">
    Text vectorization for semantic search and knowledge bases.
  </Card>

  <Card title="Voice Models" icon="microphone" href="./models/voice">
    Transcription (speech-to-text) and speech output (text-to-speech).
  </Card>
</CardGroup>

## Best Practices

* **Descriptive names**: Use descriptive names (e.g. "Azure GPT 5.2") to simplify management.
* **Configure costs**: Enter realistic token costs so that cost tracking works accurately.
* **Security**: Enable DLP protection (GDPR, Sensitive File Protection) according to your compliance requirements.
* **Testing**: Test new model configurations in a safe environment before deploying to production.

<Warning>
  Make sure that the model configuration is compatible with the selected
  credentials. Models can only be used with credentials of the corresponding
  provider type.
</Warning>

<Note>
  Fields marked with an asterisk (\*) are required and must be filled in.
</Note>