# Create AI Avatar Videos

Teams need presenter videos, dubbing, and lip-sync clips without a full studio. This skill guides Claude, Codex, and Claude Code through inference.sh CLI workflows for avatar video generation.

## Install

```bash
npx skillstore add inference-sh-skills/ai-avatar-video
```

## Metadata

- Status: approved
- Slug: inference-sh-skills-ai-avatar-video
- Skillstore revision: r1
- Version status: missing
- Tree hash: dc92e9e6c7af1f7d9f6eb816b5eccd4ba14f4b23e2905e1929852f4de9b04307
- Author: inference-sh-skills
- GitHub username: inference-sh-skills
- License: MIT
- Repository: https://github.com/inference-sh-skills/skills/tree/main/tools/video/ai-avatar-video/
- Ref: a25199bc7d6b82598536822d1738eb5d5f54025b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network
- Quality score: 79
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/inference-sh-skills-ai-avatar-video
- Manifest: https://skillstore.pages.dev/api/skills/inference-sh-skills-ai-avatar-video/manifest

## Capabilities

- Guides belt CLI workflows for P-Video-Avatar, OmniHuman, Fabric, and PixVerse.
- Creates avatar videos from a portrait and text script with built-in TTS.
- Uses audio files to drive talking head and lip-sync video models.
- Combines image generation, TTS, transcription, translation, and lip-sync steps.
- Compares model options by speed, cost, TTS support, and best use.
- Links related inference.sh video, image, speech, and platform skills.

## Use Cases

- Produce Product Presenter Clips: Create short avatar-led demos from a portrait and product script.
- Build Course Explainer Videos: Turn lesson scripts into consistent talking head videos for learners.
- Localize Existing Video: Transcribe, translate, generate new speech, and lip-sync the result.

## Prompt Templates

### Create a Basic Avatar Clip

```
Create a 20-second AI avatar video from my portrait and this script. Recommend the simplest model and settings.
```

### Use My Recorded Audio

```
Make a talking head video from my portrait and recorded audio. Choose a model that supports audio-driven lip sync.
```

### Compare Model Options

```
Compare P-Video-Avatar, OmniHuman, Fabric, and PixVerse for my project. Prioritize cost, speed, realism, and TTS support.
```

### Plan a Dubbing Pipeline

```
Design a dubbing workflow for this video. Include transcription, translation review, speech generation, lip sync, and quality checks.
```

## Limitations

- Requires the inference.sh belt CLI and an active account.
- Media and scripts are processed by external inference.sh applications.
- Output quality depends on portrait quality, audio clarity, and model limits.
- It does not verify rights, consent, or identity for avatar content.

## Best Practices

- Use consented, high-quality portraits with clear lighting and a front-facing pose.
- Start with a short preview before rendering long or high-resolution videos.
- Review scripts, voices, lip sync, and identity use before publishing.

## Anti Patterns

- Do not use photos, voices, or identities without permission.
- Do not submit private customer data or regulated data to external model inputs.
- Do not publish generated speech or lip sync without human quality review.

## Security Audit

- Audited at: 2026-07-05T17:36:00.216\+00:00
- Summary: Reviewed all 48 static findings. The command detections are Markdown fences, inline examples, or documented CLI snippets rather than runtime backtick execution. The hardcoded URLs are documentation links, placeholder media inputs, or public assets; no prompt injection or covert exfiltration intent was found.

## Stats

- Views: 73
- Downloads: 5
- Favorites: 0
- Popularity score: 0