avatar-video

Generate AI avatar videos with HeyGen's v2 API using custom avatars, voices, scripts, and scenes.

Updated Jul 16, 2026
One-click install
npx skills add https://github.com/X-manist/Cohmira --skill avatar-video-x-manist
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: avatar-video
Source: https://github.com/X-manist/Cohmira/tree/main/src/builtin-plugins/openmontage/.agents/skills/avatar-video
Command: npx skills add https://github.com/X-manist/Cohmira --skill avatar-video-x-manist

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Producing presenter-style videos normally requires cameras, actors, and editing time. This Skill lets you generate AI avatar videos programmatically through HeyGen's v2 API, with exact control over which avatar speaks, what script it says, which voice it uses, and how each scene and background is configured. ## Core Features & Use Cases - Precise Avatar and Voice Control: List and preview avatars, select voices, and use each avatar's default voice for natural lip sync. - Multi-Scene Video Generation: Build videos with multiple scenes, each with its own avatar, script, background, captions, and text overlays via POST /v2/video/generate. - Advanced Production Options: Create transparent WebM videos for compositing, talking photos from uploaded images, AI-generated photo avatars, and Remotion integration for motion graphics. - Use Case: A marketing team needs a three-scene product demo where a specific avatar reads an exact script against branded backgrounds, with captions enabled, delivered as a 1080p landscape MP4. ## Quick Start Ask the AI to generate a HeyGen avatar video where a chosen avatar reads your exact script, then poll the video status until the download URL is ready.

Frequently Asked Questions about avatar-video

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

FAQPage Schema
How do I generate an AI avatar video with HeyGen API?

Call POST /v2/video/generate with a video_inputs array specifying the avatar_id, voice_id, input_text script, and background per scene. Authenticate with the X-Api-Key header, then poll GET /v2/videos/{video_id} until the status is completed.

How do I choose the right voice for a HeyGen avatar?

Fetch the avatar's details via GET /v2/avatar/{avatar_id}/details and use its default_voice_id, which is pre-matched for natural lip sync. If no default exists, list voices with GET /v2/voices and match the voice gender to the avatar.

Can HeyGen create videos with a transparent background?

Yes, use the /v1/video.webm endpoint to generate WebM videos with transparent backgrounds for compositing. WebM supports only normal and closeUp avatar styles; apply circular masks in CSS if needed.

How do I create a talking photo avatar from an image?

Upload the image to upload.heygen.com/v1/asset to get an image_key, then create a photo avatar group via POST /v2/photo_avatar/avatar_group/create. Once processing completes, use the returned id as talking_photo_id in video generation.

Why is my HeyGen avatar video jittery in Remotion?

Jitter occurs when using Remotion's basic Video component, which relies on the browser's non-frame-accurate decoder. Switch to OffthreadVideo from the core remotion package, which extracts frames via FFmpeg for accurate rendering.

How long does HeyGen video generation take?

Generation typically takes 5-15 minutes, sometimes longer depending on length and resolution. Poll the video status endpoint with generous timeouts, and use test mode with watermarked output during development to avoid consuming credits.