Audio Transcriptions

Name: Chuizi AI
Price: 5 USD

POST /v1/audio/transcriptions

The transcription endpoint forwards multipart audio to the selected upstream model and returns JSON. Use the default JSON response for stable integrations; do not rely on text, srt, or vtt output through the gateway.

Request

terminal

bash

curl -X POST https://api.chuizi.ai/v1/audio/transcriptions \
  -H "Authorization: Bearer ck-your-key" \
  -F file=@recording.mp3 \
  -F model="openai/whisper-1" \
  -F language="en"

Parameter	Required	Description
`file`	Yes	Audio file
`model`	No	Defaults to `openai/whisper-1`; STT models from the catalog are also supported
`language`	No	ISO 639-1 language code
`prompt`	No	Optional context prompt
`response_format`	No	Omit it, or use `json` / `verbose_json`

Response

config.json

json

{
  "text": "Hello, this is a test recording."
}