HINOW model
hinow/hivox
HiVox
Audio transcription into text, from an uploaded file or a URL.
Best for: Meetings, calls, in-app audio, and video
- Meetings and calls — turn the recording into text to summarize or index later.
- Audio from your application — the voice message the user records instead of typing.
- Customer service — transcribe the call to classify the subject and measure quality.
- Video — the audio track becomes captions, a summary, or searchable material.
- Input to a second step — transcribe, then send the text to a language model to extract what matters.
- 1
Choose how to send it
A file in
multipart/form-dataunderfile, or the address of an audio file the API can reach. - 2
Receive the text
The response carries the duration and the transcript — a short body, with no per-segment timestamps.
- 3
Chain the next step
Send the text to Chat Completions when you need a summary, a classification, or structured fields.
curl https://api.hinow.ai/v1/audio/transcriptions \
-H "Authorization: Bearer $HINOW_API_KEY" \
-F "model=hinow/hivox" \
-F "file=@reuniao.mp3"Billing is per minute
Unlike the language models, there is no charge per token here: the price follows the duration of the audio. Trimming silence and irrelevant stretches before sending lowers the bill.
See the full guide
Accepted formats, sending by file and by URL, measured times, and the list of errors.