skills.rest
inference-sh
inference-shOFFICIAL
Explore all skills by this author →
WebsiteGitHub

p-video-avatar

Generate talking head avatar videos from portrait images using Pruna P-Video-Avatar via the inference.sh CLI.

723|101|Updated Jan 31, 2026
Search Tags:#text-to-speech#video generation#inference.sh#lipsync#talking head#avatar video#ai presenter
One-click install
npx skills add https://github.com/inference-sh/skills --skill p-video-avatar
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: p-video-avatar
Source: https://github.com/inference-sh/skills/tree/main/tools/video/p-video-avatar
Command: npx skills add https://github.com/inference-sh/skills --skill p-video-avatar

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Creating presenter-style videos traditionally requires cameras, actors, and editing time. This Skill turns a single portrait image into a speaking avatar video with built-in text-to-speech, removing the need for filming or separate voiceover production.

Core Features & Use Cases

  • Text-to-Avatar Generation: Provide a portrait image and a voice script to produce a lip-synced talking head video at 720p or 1080p.
  • Audio-Driven Avatars: Supply your own audio file to drive the avatar instead of using the built-in TTS with 30 voices across 10 languages.
  • Style and Behavior Control: Use voice_prompt and video_prompt parameters to control tone, pacing, emotion, backgrounds, and body language.
  • Use Case: A marketing team generates a product demo video by creating a portrait with pruna/p-image, then animating it with a scripted voiceover in Spanish and English versions for localized campaigns.

Quick Start

Ask the AI to generate a talking avatar video from a portrait image URL with a short welcome script using the p-video-avatar skill.

Frequently Asked Questions about p-video-avatar

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a talking avatar video from a photo?▼

Run belt app run pruna/p-video-avatar with an image URL and a voice_script parameter. The model generates a lip-synced video of the person in the portrait speaking your script, with output at 720p by default.

Can I use my own audio instead of text-to-speech for avatar videos?▼

Yes, pass an audio file URL in the audio parameter instead of voice_script. When both are provided, the audio input takes priority and drives the avatar's lip movements directly.

What languages and voices does P-Video-Avatar support?▼

It supports 10 languages including English, Spanish, French, German, Japanese, Korean, and Hindi. There are 30 built-in voices split between female and male options, selectable via the voice parameter.

How much does it cost to generate an avatar video?▼

Pricing is per second of output video: $0.025 per second at 720p and $0.045 per second at 1080p. A 30-second 720p video costs approximately $0.75.

What image works best for generating avatar videos?▼

Use high-quality front-facing portrait photos with good lighting in jpg, jpeg, png, or webp format. The output video matches the input image aspect ratio, and 9:16 portraits work well for vertical videos.

How do I control the avatar's tone and background?▼

Use voice_prompt to control speaking tone, pacing, and emotion, and video_prompt to control backgrounds and body language. For example, set video_prompt to describe a stage presentation with dramatic lighting.

Related Skills

Explore More Content & Communication Skills→
inference-shinference-sh
688

p-image

Generate and edit images with Pruna P-Image models via inference.sh CLI.

Official
Intermediate
refly-airefly-ai
199

video-creator

Generates short AI-powered videos and scripts for social media platforms.

Official
Intermediate
GunjanGrungeGunjanGrunge
3

video-pipeline

Orchestrate a multi-stage video pipeline from script to MP4 with EC2 GPU instances.

Community
Advanced
inference-shinference-sh
688

ai-avatar-video

Generates AI avatar and talking head videos via the inference.sh CLI.

Official
Intermediate