Talking head API: one photo in, a lip-synced video out
One POST with a photo URL and either text or an audio URL returns a job. A signed webhook tells you when the MP4 is ready. It costs $0.20 per minute at 720p, which is the lowest published price we found for an API that takes your own photo and your own audio.
What it costs, next to published prices
| API | Price and limits |
|---|---|
| Stork Standard | $0.20 per minute at 720p. From $0.10 at 480p. Any length. |
| Stork Pro | $0.40 per minute at 720p. Sharper and steadier. |
| D-ID API (annual plans) | About $0.69 to $0.78 per minute, our arithmetic from plan prices. Subscription. |
| HeyGen API, Avatar III photo avatar | $0.99 per minute, plus $1.32 to create each photo avatar. |
| HeyGen API, Avatar IV photo avatar | $2.31 per minute. |
| Hedra Character-3 | $3.00 per minute at 720p. Up to 10 minutes per video. |
| OmniHuman 1.5 on fal | $9.60 per minute at 720p. Up to 60 seconds per call. Animates body and hands. |
Other companies' prices were read on their own pages on November 13, 2026 and may have changed. Sources are listed below. Stork's prices are current.
Use Stork when
- Price per minute decides the project: narration, lessons, news recaps, support videos at volume.
- You need long videos in one call.
- You want to start free and add a card later.
Use something else when
- You need body and hand motion. Models such as OmniHuman animate the whole upper body; Stork animates the face and head only.
- You need a real-time, streaming avatar today.
One request
Send text and a voice_id instead of audio_url to use a built-in voice, or "tier": "free" for a watermarked test video.
curl -X POST https://mcp.stork.ai/api/v1/talking-photo \
-H "Authorization: Bearer sk_stork_YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{
"image_url": "https://www.stork.ai/samples/talking-photo/portrait.jpg",
"audio_url": "https://example.com/speech.mp3",
"aspect_ratio": "16:9",
"webhook_url": "https://example.com/hooks/stork"
}'The response is a job with a status URL. When the MP4 is ready, your webhook gets a signed POST. Idempotency keys, an estimate endpoint, an OpenAPI file and an MCP server are in the API docs.
Questions
- What is a talking head API?
- An API that takes a still image of a face and speech (text to be read, or an audio file) and returns a video in which the face speaks in sync with the audio.
- Is it really the cheapest?
- It is the lowest published price we found on November 13, 2026 among APIs that accept your own photo and your own audio. We list every service we could verify, with links, in the pricing post below. Prices change, so check the sources.
- How long can a video be?
- There is no limit on paid videos. We have rendered a 33-minute video in one call.
- Which languages does it support?
- With your own audio, any language: the lips follow the sound. The built-in text-to-speech voices cover 13 languages.
Sources
- Full price table and our five-API test
- HeyGen API pricing explained
- D-ID API pricing
- Hedra Character-3
- OmniHuman 1.5 on fal
More comparisons: HeyGen API alternative · D-ID API alternative
