Introduction
HeyGen Video Translate changes the language of speech in a short video and syncs the lips to the translation. HeyGen says it supports 175+ languages and dialects with realistic AI voices, and that voice cloning carries the speaker’s tone across languages. This page offers 178 target options.
Get consent before sending footage: use a recording of yourself or people who agreed to this use. Translation makes them appear to say words in another language in a voice like their own. The first run on the model page requires acceptance of a consent notice.
Provide a spoken MP4 or WebM no longer than 120 seconds. Set target_language and, if needed, generation_mode; there is no prompt to write. English and Precision are the defaults.
Key features & highlights
- Lip movement changes with the speech. In our tests, source and translated frames at matching moments had different mouth shapes, while faces, clothing and the scene stayed the same.
- Choose a regional target. The 178 options include Spanish (Mexico), French (Canada) and Chinese (Cantonese, Traditional), alongside base languages.
- Two modes set the tradeoff. Precision aims for higher-quality lip sync, while Speed aims for a faster result and uses half the per-second rate.
- English accent choices are available. Targets include “English - Your Accent” and “English - American Accent.”
- Frame size is retained in our tests. We received MP4 with AAC audio; a 1280x720 input returned at the same size.
- Charges use the original duration. A longer translated result does not add billed source seconds; failed or refused tasks incur no charge.
Request parameters
| Parameter | Required | Description |
|---|---|---|
video_url | Yes | Required spoken video, limited to 120 seconds. The model page accepts MP4 and WebM. at least 1 characters |
target_language | No | Defaults to English. Select among 178 base-language and regional-dialect options, including two English accent targets. |
generation_mode | No | Use precision, the default, for its higher-quality lip-sync aim. Use speed for its faster-result aim at half the Precision rate; completion time is not guaranteed. |
Pricing
Charges follow the source video’s duration, rounded upward to whole seconds, with a minimum of 1 second and a maximum of 120 seconds. Precision is billed at twice the Speed rate. Language selection does not affect the rate. Failed and refused tasks incur no charge, including clips refused for exceeding 120 seconds.
| Specification | Price |
|---|---|
| $0.048 / second |
| $0.096 / second |
Use cases
- Course welcome in Spanish: translate your own recorded introduction so Spanish-speaking students can hear the welcome with matching lip movement.
- French product demonstration: get your presenter’s consent, then translate a short walkthrough for French-speaking customers.
- Japanese creator announcement: adapt your spoken update for Japanese-speaking viewers using the same source recording.
Prompt guide & pro tips
- Check permission first. Obtain agreement from everyone who will appear to speak translated lines; do not use a celebrity, public figure or acquaintance without permission.
- Check the recording before submission. Listen for audible speech and trim the source to 120 seconds or less; longer videos are refused before a charge.
- Select the target deliberately. Check whether a base language or regional dialect suits your viewers, rather than leaving the English default by accident.
- Choose Precision for its lip-sync quality aim. It is the default and carries twice the Speed rate, so include that difference in your decision.
- Choose Speed when its rate and timing aim suit the task. It aims to finish sooner at half the per-second rate; actual completion time can vary.
- Allow space for longer translated speech. In our tests, an 8.00-second recording returned as 8.68 seconds in Spanish; check the finished duration before editing it into a sequence.
- Save the returned video within 14 days. Generated videos are deleted after that retention period.
- Use the available controls. There is no prompt, source-language field or subtitle option on this page; choose the clip, target language and mode instead.
API quick start
# Generate the idempotency key once; reuse it when retrying after a lost response.
export SPICY_IDEMPOTENCY_KEY="${SPICY_IDEMPOTENCY_KEY:-$(uuidgen)}"
curl -X POST https://api.spicyapi.ai/api/v1/jobs/createTask \
-H "Authorization: Bearer $SPICY_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: $SPICY_IDEMPOTENCY_KEY" \
-d '{
"model": "heygen/video-translate/video-translate",
"input": {
"generation_mode": "precision",
"target_language": "English",
"video_url": "https://example.com/reference.mp4"
}
}'
# Copy data.taskId from the response into TASK_ID, then poll the status.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.spicyapi.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
-H "Authorization: Bearer $SPICY_API_KEY"
See the API reference for authentication, the complete schema, polling, and error handling.

