Stream Speech

Generates speech as Server-Sent Events. Audio events contain ordered base64-encoded standalone WAV parts; alignment events are included only when with_timestamps is true. A clean stream ends with data: [DONE]. The voice identified by voice_id must be an ACTIVE professional voice. Rate limit: 30 requests per minute per workspace member.

POST /v3/models/audio/tts/stream

Stream Speech

cURL

curl --request POST \
  --url https://api.heygen.com/v3/models/audio/tts/stream \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: <api-key>' \
  --data '
{
  "voice_id": "<string>",
  "text": "<string>",
  "language": "<string>",
  "seed": 2147483647,
  "with_timestamps": false
}
'

200

Example

{
  "type": "audio",
  "part_index": 1,
  "audio": "<string>"
}

Authorizations

ApiKeyAuthBearerAuth

x-api-key · string · header · required

HeyGen API key. Obtain from your HeyGen dashboard.

Body

application/json

voice_id · string · required

Identifier of the voice used for speech synthesis.

Minimum string length: 1

text · string · required

Plain text to synthesize. SSML and break tags are not supported.

Required string length: 1 - 5000

language · string · required

Language code used for speech synthesis, such as en.

Minimum string length: 1

seed · integer | null

Optional best-effort deterministic generation seed.

Required range: 0 <= x <= 4294967295

with_timestamps · boolean · default: false

Whether to include word-timestamp events in the response stream.

Response

200

text/event-stream

Server-Sent Events stream. Each data: frame contains an audio, alignment, or terminal error event; a clean stream terminates with data: [DONE].

SpeechAudioEvent

SpeechAlignmentEvent

SpeechErrorEvent

One JSON event carried by the speech SSE stream.

type · string · required

Event type.

Allowed value: "audio"

part_index · integer · required

Zero-based sequence number of this audio part.

Required range: x >= 0

audio · string · required

Base64-encoded standalone WAV audio part.