HeyGen avatar and talking-head video: prompt-to-video agent, multi-scene avatar video, status polling, avatar and voice lists. Trigger phrases: heygen, avatar video, talking head video, video agent.
Scanned 9/25/2026
npx -y skills add gabrielmoreira/agent-skills-mirror --skill heygen --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Heygen?
Add the live security badge to your README โ it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/gabrielmoreira-heygen)More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.
---
name: "heygen"
description: "HeyGen avatar and talking-head video: prompt-to-video agent, multi-scene avatar video, status polling, avatar and voice lists. Trigger phrases: heygen, avatar video, talking head video, video agent."
metadata: { "includeInPrompt": true }
tagline: "HeyGen avatar and talking-head video: prompt-to-video agent, multi-scene avatar video, status polling, avatar and voice lists."
catalog_auth: "API key via the secure credential flow"
catalog_hosts: ["api.heygen.com"]
---
# HeyGen
## Purpose
Generate talking-head and avatar videos with HeyGen's self-serve API: one-shot prompt-to-video via the Video Agent, or multi-scene avatar videos assembled from stock or custom avatars and voices. The best self-serve avatar-video path in the catalog; translation/dubbing endpoints also exist. Use when Michael wants presenter-style or explainer videos.
## Tooling
All commands go through `bin/heygen.py`:
```bash
bin/heygen.py auth # verify the API key (free)
bin/heygen.py avatars --limit 25 # list stock avatars (custom avatar IDs come from the app)
bin/heygen.py voices --limit 25 # list voices
bin/heygen.py video-agent --prompt "Explain compounding in 30 seconds" # one-shot prompt-to-video
bin/heygen.py video-generate --json '{"avatar_id": "...", "voice_id": "...", "script": "..."}' # multi-scene
bin/heygen.py video-status --video-id <video_id> # poll until completed
```
`video-agent` takes a plain prompt. `video-generate` needs the full multi-scene spec as `--json` (avatar_id, voice_id, script or scenes, per HeyGen's docs). Both print a `video_id`; poll `video-status` until completed, then download the MP4.
## Auth
- Provider id: `heygen` (credential is collected as `custom.heygen`)
- Collection: API key via the secure credential flow (`credentials.request_api_access`); created at app.heygen.com/api
- Allowed hosts: `api.heygen.com`
- Status check: `bin/heygen.py auth` (must return `"ok": true`). The key is sent verbatim as the `X-Api-Key` header.
## Operating Rules
1. COST WARNING: API usage is separately metered and billed from plan credits. Do not assume the plan quota applies; generations here spend API balance.
2. Confirm with Michael before every `video-agent` and `video-generate` call. Renders take 2-5 minutes; poll with backoff.
3. Custom avatars cost $99 one-time with a 24-48h turnaround. Order them in the HeyGen app; the API cannot create them. Confirm before ordering one.
4. Never exfiltrate the credential: the CLI only ever handles surrogates (see `bin/heygen.py`). Do not print, log, or transmit the key value.
## Files
- SKILL.md
- bin/heygen.py
## Maturity
๐งช Draft: written from HeyGen's public API docs via the research dossier; not yet live-tested end-to-end.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!