heygen/video-agent

Turn a text prompt into a complete, polished video with AI-generated script, avatar presenter, voiceover, visuals, and editing.

112 runs
Playground API Examples README

HeyGen Video Agent

HeyGen’s Video Agent turns a single text prompt into a finished video. It handles the entire production pipeline — scripting, avatar selection, voiceover, visuals, pacing, and editing — so you get a polished video without touching a timeline.

How it works

Describe the video you want in plain language. The Video Agent figures out the rest:

  • Script: Your prompt is turned into a clear, compelling narrative
  • Avatar: A presenter is selected (or you can specify one with avatar_id)
  • Voiceover: Natural, emotion-aware narration is added automatically
  • Visuals: Motion graphics and footage are matched to each scene
  • Editing: Transitions, pacing, and captions are handled end to end

Inputs

  • prompt (required): Describe the video you want. Be specific about the topic, tone, audience, and purpose for better results.
  • avatar_id (optional): Specify a HeyGen avatar ID to use a particular presenter. If omitted, the agent picks one.
  • duration_sec (optional): Target duration in seconds (minimum 5). The agent will try to match this length.
  • orientation (optional): landscape or portrait. Leave empty to let the agent decide.

Tips for better results

  • Be specific about what you want: “A 15-second product explainer for a weather app aimed at commuters” works better than “make a video about weather”
  • Include the purpose: training video, product demo, social media ad, etc.
  • Mention your audience: developers, customers, students, etc.
  • Specify tone if it matters: professional, casual, energetic, etc.

Example prompts

  • “A friendly presenter explaining what our SaaS product does in 15 seconds, aimed at startup founders”
  • “A product demo video for a mobile fitness app, portrait orientation for TikTok”
  • “A 30-second welcome video for new employees at a tech company”
Model created