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

# Overview

> Understand Tavus Faces, learn how to create your own using a video or an image, or explore ready-to-use stock Faces.

## What is a Face?

A **Face** is a photorealistic likeness in CVI - it defines the on-screen appearance and voice. Faces are trained with **Phoenix** and paired with a **[PAL](/sections/conversational-video-interface/pal/overview)** that defines behavior, knowledge, and pipeline settings.

With Tavus, you can create a face that looks like you - or use a stock face for instant prototyping. Faces capture natural expressions, tone, and presence so interactions feel human and engaging.

Faces are generated using our Phoenix-4 model, built on advanced research in real-time human rendering and trained on large-scale conversational data to model natural expression, emotion, and behavior.

## Custom Faces

You can create a face in one of two ways:

* **<u>Upload a training video</u>** – Record a 1-minute video with **30 seconds of speaking followed by 30 seconds of listening**. This method provides the highest quality and most control. With this short input, Tavus can reproduce appearance with high fidelity, voice, expressions, and accurate lip-sync. See [Training from a Video](/sections/faces/train-with-a-video).
* **<u>Upload an image</u>** – Upload a photo of a person or character, and Tavus will automatically generate the training data and create the face for you. This path is faster to set up but lower fidelity than video. See [Training from an Image](/sections/faces/train-with-an-image).

<Note>
  Creating a custom face is **only available** on the Starter, Growth, and Enterprise plans.
</Note>

## Stock Faces

Alternatively, Tavus offers 100+ ready-to-use [stock Faces](/sections/faces/stock-faces) recorded using real actors - great for quick prototyping, demos, or teams without access to custom training footage or images.

## Key Features

<CardGroup cols={3}>
  <Card title="Realistic Face Cloning" icon="face-smile">
    Custom faces capture a person's look, expressions, and speaking style.
  </Card>

  <Card title="Emotional Control" icon="arrows-repeat">
    Phoenix-4 faces express emotions through lifelike facial expressions while speaking and listening.
  </Card>

  <Card title="Full Motion Control" icon="head-side">
    Smooth transitions between speaking and listening with fully generated, natural head motion.
  </Card>
</CardGroup>

<Note>
  Emotional expression controls in conversational video apply only to **Phoenix-4**; see [Emotion Control with Phoenix-4](/sections/conversational-video-interface/quickstart/emotional-expression).
</Note>

## Platform Policies

By using PAL Maker and the Tavus APIs, you agree to comply with Tavus' [Terms of Service](https://www.tavus.io/terms-of-service) and [Acceptable Use Policy](https://www.tavus.io/acceptable-use-policy). You are responsible for ensuring you have the necessary rights and permissions to upload and use any training data, including videos, images, voices, and likenesses.

## Getting Started

You can create a face using either the PAL Maker or the API, depending on your workflow.

* For a guided face creation flow with upload checks and inline validation, use the [PAL Maker](https://maker.tavus.io/dev/faces/create).
* For API flows, use [Video to Face](/sections/faces/video-to-face-quickstart) or [Image to Face](/sections/faces/image-to-face-quickstart). For recording rules, footage requirements, and best practices, see [Which training path?](/sections/faces/which-training-path)