Audio

Audio APIs include speech generation, transcription and translation. Speech generation uses a JSON request body; transcription and translation usually use multipart form data to upload the audio file.

Endpoints

Method	Path	Description
`POST`	`/v1/audio/speech`	Text to speech
`POST`	`/v1/audio/transcriptions`	Audio transcription
`POST`	`/v1/audio/translations`	Audio translation

Request Fields

Field	Type	Applies To	Description
`model`	string	All	Audio model ID, required
`input`	string	speech	Text to synthesize
`voice`	string	speech	Voice ID
`instructions`	string	speech	Voice style or extra instructions
`response_format`	string	All	Response format. Transcription and translation default to `json`
`speed`	number	speech	Speech speed
`stream_format`	string	speech	Set to `sse` to enable streaming when the model supports it
`metadata`	object	speech	Pass-through metadata
`file`	file	transcriptions/translations	Audio file in multipart form data, required
`language`	any	extension	Language option supported by some channels

Speech Generation

curl https://api.tensoraxis.ai/v1/audio/speech \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $TENSORAXIS_API_KEY" \
  -d '{
    "model": "tts-1",
    "voice": "alloy",
    "input": "Hello from TENSORAXIS.",
    "response_format": "mp3"
  }' \
  --output speech.mp3

Transcription

curl https://api.tensoraxis.ai/v1/audio/transcriptions \
  -H "Authorization: Bearer $TENSORAXIS_API_KEY" \
  -F model="whisper-1" \
  -F file="@audio.mp3" \
  -F response_format="json"

Translation

curl https://api.tensoraxis.ai/v1/audio/translations \
  -H "Authorization: Bearer $TENSORAXIS_API_KEY" \
  -F model="whisper-1" \
  -F file="@audio.mp3" \
  -F response_format="json"

Response Shape

JSON transcription and translation responses usually include text:

{
  "text": "Transcribed text..."
}

Speech generation returns binary audio directly. The response Content-Type depends on response_format and the upstream model.