Actions8
Video → Create Video Agent Video
AI-generatedSummary
Submit a natural-language prompt to HeyGen Video Agent to generate a video. Optionally specify an avatar, voice, orientation, brand kit, style, callback URL, or incognito mode; return the agent session and video status.
Inputs
- Prompt (required) — Describe the video to create. The node requires non-empty text and rejects prompts over 10,000 characters.
- Avatar Type — Optionally filters the avatar picker to studio, digital twin, or photo avatars; all types are shown by default.
- Avatar — Optionally select an avatar look to send as the requested avatar ID. If omitted, Video Agent chooses.
- Voice — Optionally select a voice to send as the requested voice ID. If omitted, Video Agent chooses.
- Orientation — Choose automatic, landscape, or portrait. Automatic is the default and omits orientation from the request.
- Options — Optional Brand Kit ID, Callback URL, Incognito Mode, and Style ID. Callback URL must be HTTP or HTTPS; Incognito Mode is sent only when enabled and disables Video Agent memory injection and extraction for the session.
Output shape
A single item per input item, containing session_id, video_id, and status.
video_id is null when HeyGen does not return one; status is processing when a video ID is present and queued otherwise. The operation submits generation and returns without waiting for rendering to finish.
Examples
Example 1: prompt
Provide a concise script prompt and leave avatar, voice, orientation, and optional settings empty. The agent will select defaults automatically and return session_id with video_id/status if available.
Example 2: portrait_brand
Select a specific avatar and voice, set orientation to portrait, and provide a Brand Kit ID and Callback URL in Options. The returned status is processing or queued based on whether HeyGen returned a video ID.