# Overview

Animate portraits and create talking head videos on CLORE.AI GPUs.

## Available Guides

| Tool                                                                    | Use Case           |
| ----------------------------------------------------------------------- | ------------------ |
| [LivePortrait](https://docs.clore.ai/guides/talking-heads/liveportrait) | Portrait animation |
| [SadTalker](https://docs.clore.ai/guides/talking-heads/sadtalker)       | Audio-driven faces |
| [Wav2Lip](https://docs.clore.ai/guides/talking-heads/wav2lip)           | Lip sync           |

## Comparison

| Tool         | Quality | Speed  | Lip Sync |
| ------------ | ------- | ------ | -------- |
| LivePortrait | Best    | Fast   | Good     |
| SadTalker    | Good    | Medium | Great    |
| Wav2Lip      | Medium  | Fast   | Best     |

## Workflow

1. Generate/obtain portrait image
2. Generate/record audio ([TTS guides](https://docs.clore.ai/guides/audio-and-voice/audio-voice))
3. Animate with chosen tool
4. Post-process if needed

## GPU Requirements

6-8GB VRAM minimum for all tools.

## Related Guides

* [Audio & Voice](https://docs.clore.ai/guides/audio-and-voice/audio-voice)
* [Face Identity](https://docs.clore.ai/guides/face-and-identity/face-identity)