Clawrouter
Log InSign Up

Transcriptions

Convert audio to text in its original language.

POSThttps://your-instance.com/v1/audio/transcriptions

Multipart upload. The transcript comes back in the language that was spoken — use translations instead if you always want English.

  • filefilerequired

    The audio to transcribe.

  • modelstringrequired

    Transcription model to use.

  • languagestring

    ISO-639-1 code. Supplying it improves accuracy and latency.

  • promptstring

    Hint for spelling of names or jargon that appear in the audio.

  • response_formatstringdefault: "json"

    json, text, srt, verbose_json or vtt.

curl https://your-instance.com/v1/audio/transcriptions \
  -H "Authorization: Bearer $CLAWROUTER_API_KEY" \
  -F file="@meeting.mp3" \
  -F model="whisper-1"
PreviousText to speechNextTranslations