Agent Skills: heygen

Troubleshooting

If requests fail, run zero doctor check-connector --env-name HEYGEN_TOKEN or zero doctor check-connector --url https://api.heygen.com/v2/avatars --method GET

How to Use

All examples below assume you have HEYGEN_TOKEN set.

The base URL for the HeyGen API is:

• v1 endpoints: https://api.heygen.com/v1
• v2 endpoints: https://api.heygen.com/v2

1. List Available Avatars

Get all avatars available to your account:

curl -s -X GET "https://api.heygen.com/v2/avatars" --header "x-api-key: $HEYGEN_TOKEN" | jq '.data.avatars[] | {avatar_id, avatar_name, gender}'

Each avatar has an avatar_id needed for video generation.

2. Get Avatar Details

Retrieve detailed information about a specific avatar. Replace <avatar_id> with an actual avatar ID:

curl -s -X GET "https://api.heygen.com/v2/avatar/<avatar_id>/details" --header "x-api-key: $HEYGEN_TOKEN" | jq .data

3. List Available Voices

Get all AI voices for video narration:

curl -s -X GET "https://api.heygen.com/v2/voices" --header "x-api-key: $HEYGEN_TOKEN" | jq '.data.voices[] | {voice_id, name, language, gender}'

Voice properties include:

• voice_id: Unique identifier for use in video generation
• language: Primary language (e.g., "English", "Multilingual")
• gender: "female", "male", or "unknown"
• emotion_support: Whether the voice supports emotion variations

4. Create an Avatar Video

Generate a video with an AI avatar presenter. Replace <avatar_id> and <voice_id> with actual IDs from the list endpoints.

Write to /tmp/heygen_request.json:

{
  "video_inputs": [
    {
      "character": {
        "type": "avatar",
        "avatar_id": "<avatar_id>",
        "avatar_style": "normal"
      },
      "voice": {
        "type": "text",
        "input_text": "Hello! Welcome to our product demo. Let me walk you through the key features.",
        "voice_id": "<voice_id>",
        "speed": 1.0
      },
      "background": {
        "type": "color",
        "value": "#FFFFFF"
      }
    }
  ],
  "dimension": {
    "width": 1920,
    "height": 1080
  },
  "title": "Product Demo Video"
}

Then run:

curl -s -X POST "https://api.heygen.com/v2/video/generate" --header "x-api-key: $HEYGEN_TOKEN" --header "Content-Type: application/json" -d @/tmp/heygen_request.json | jq .

The response contains a video_id to track the generation progress.

5. Check Video Status

Poll for video generation status using the video_id from the generate response. Replace <video_id> with the actual video ID:

curl -s -X GET "https://api.heygen.com/v1/video_status.get?video_id=<video_id>" --header "x-api-key: $HEYGEN_TOKEN" | jq '{status: .data.status, video_url: .data.video_url, duration: .data.duration}'

Status values:

• pending: Video is queued
• processing: Video is being rendered
• completed: Video is ready (includes video_url)
• failed: Generation failed (includes error details)

6. List Videos

Retrieve all videos associated with your account:

curl -s -X GET "https://api.heygen.com/v1/video.list" --header "x-api-key: $HEYGEN_TOKEN" | jq .

7. Delete a Video

Remove a video from your account. Replace <video_id> with the actual video ID:

curl -s -X DELETE "https://api.heygen.com/v1/video.delete" --header "x-api-key: $HEYGEN_TOKEN" --header "Content-Type: application/json" -d '{"video_id": "<video_id>"}' | jq .

8. List Templates

Get all video templates created in your account:

curl -s -X GET "https://api.heygen.com/v2/templates" --header "x-api-key: $HEYGEN_TOKEN" | jq .data

9. Get Template Details

Retrieve a template configuration and its variables. Replace <template_id> with the actual template ID:

curl -s -X GET "https://api.heygen.com/v2/template/<template_id>" --header "x-api-key: $HEYGEN_TOKEN" | jq .data

10. Generate Video from Template

Create a video using a template with variable substitution. Replace <template_id> and variable keys with actual values from the template details.

Write to /tmp/heygen_request.json:

{
  "caption": false,
  "title": "Video from Template",
  "variables": {
    "script": {
      "name": "script",
      "type": "text",
      "properties": {
        "content": "Your custom script text goes here."
      }
    }
  }
}

Then run:

curl -s -X POST "https://api.heygen.com/v2/template/<template_id>/generate" --header "x-api-key: $HEYGEN_TOKEN" --header "Content-Type: application/json" -d @/tmp/heygen_request.json | jq .

11. Translate a Video

Translate an existing video into another language. Replace <video_url> and <language> with actual values:

Write to /tmp/heygen_request.json:

{
  "video_url": "<video_url>",
  "output_language": "<language>",
  "title": "Translated Video"
}

Then run:

curl -s -X POST "https://api.heygen.com/v2/video_translate" --header "x-api-key: $HEYGEN_TOKEN" --header "Content-Type: application/json" -d @/tmp/heygen_request.json | jq .

12. List Supported Translation Languages

Get all languages available for video translation:

curl -s -X GET "https://api.heygen.com/v2/video_translate/target_languages" --header "x-api-key: $HEYGEN_TOKEN" | jq .data

13. Share a Video

Generate a public sharing URL for a video. Replace <video_id> with the actual video ID:

curl -s -X POST "https://api.heygen.com/v1/video/share" --header "x-api-key: $HEYGEN_TOKEN" --header "Content-Type: application/json" -d '{"video_id": "<video_id>"}' | jq .

Guidelines

1. Poll for completion: Video generation is asynchronous. After calling the generate endpoint, poll video_status.get until status is completed or failed
2. Download promptly: Video and thumbnail URLs expire after 7 days
3. Use templates for consistency: Create templates in the HeyGen dashboard and use the template generate endpoint for repeatable video production
4. Choose appropriate avatars: List avatars first to find one matching your use case, then get details for supported styles
5. Monitor credits: Video generation consumes API credits based on video duration
6. Use JSON files for complex bodies: Write request bodies to /tmp/heygen_request.json to avoid shell quoting issues