'timestamp granularity', new featured added to AudioAPI Feb 9th not accessible

[fanpeeps]

Confirm this is an issue with the Python library and not an underlying OpenAI API

• This is an issue with the Python library

Describe the bug

audio.transcriptions.create() parameter 'timestamp_granularities' not recognized.

As per API changelog, parameter added to API Feb 9th:
https://platform.openai.com/docs/changelog

To Reproduce

passing timestamp_granularities=["word"] to audio.transcriptions.create() results in

TypeError: Transcriptions.create() got an unexpected keyword argument 'timestamp_granularities'

Code snippets

---> 19 transcript = client.audio.transcriptions.create( #openai.audio.transcriptions.create(
     20   file=wavfile,
     21   model="whisper-1",
     22   response_format="verbose_json",
     23   timestamp_granularities=["word"]
     24 )

TypeError: Transcriptions.create() got an unexpected keyword argument 'timestamp_granularities'

OS

All

Python version

3.11.4

Library version

openai-1.12.0

[fanpeeps]

Documented parameter under Timestamps heading on this page:

https://platform.openai.com/docs/guides/speech-to-text/supported-languages

[rattrayalex]

Thanks for raising this. I'm checking with the appropriate team what the correct course of action is.

[rattrayalex]

ah, this is in the latest release; you just need to upgrade your package.