Text to speech

Call the historical MiniMax t2a_v2 capability through the unified /v1/audio/speech endpoint.

Stable models currently exposed by the gateway: speech-2.8-hd, speech-2.8-turbo, speech-2.6-hd, speech-2.6-turbo, speech-02-hd, and speech-02-turbo.

POST /v1/audio/speech

cURL

basic

curl --request POST \
  --url https://api.powertokens.ai/v1/audio/speech \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data @- <<EOF
{
  "model": "speech-2.8-hd",
  "input": "Summarize today's system status in a natural tone.",
  "voice": "Chinese (Mandarin)_Lyrical_Voice",
  "response_format": "mp3",
  "metadata": {
    "output_format": "url"
  }
}
EOF

200

"<string>"

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Body

application/json

model · enum<string> · required

Available options: speech-2.8-hd, speech-2.8-turbo, speech-2.6-hd, speech-2.6-turbo, speech-02-hd, speech-02-turbo

input · string · required

Input text to synthesize.

voice · string · required

Voice name. Mapped to upstream voice_setting.voice_id.

speed · number

Speaking rate. Mapped to upstream voice_setting.speed.

response_format · string

Target audio encoding format, mapped to upstream audio_setting.format. Common values include mp3, wav, and pcm.

stream_format · string

Streaming trigger field. Any non-empty value enables upstream stream=true; the current gateway only relies on its non-empty semantics.

metadata · object

MiniMax extension container. The only stable field currently documented here is output_format.

Show child attributes

Response

200

text/event-stream

Streaming success. Returns MiniMax SSE payloads. For internal settlement, the gateway reads extra_info.usage_characters as the unified input audio character count.

SSE payload. Each data: event can be parsed as a MiniMax response chunk containing data.audio, trace_id, and base_resp.