ai-avatar-video

Turn portraits and scripts into talking-head videos.

video generation

What it does

Creates AI avatar, talking-head, lipsync, dubbing, and virtual-presenter videos.

When to use it

Use for AI presenters, explainers, virtual influencers, dubbing, marketing, UGC advertisements, and game characters.

Give it an avatar-video request; it runs an appropriate inference.sh workflow and returns the generated media result.

This skill

speech.json, transcript.json, and new_speech.json

Write

Overwrites local workflow JSON files (irreversible)

inference.sh

Write

Sends media and prompts to inference.sh

inference.sh apps

Execute

Starts hosted inference jobs

npx skills add belt-sh/cli

Execute

Installs the belt CLI skill

Requires the inference.sh CLI named belt.

Requires signing in to inference.sh through belt.

The skill directs the agent to install the belt CLI skill needed for its workflows.