- Joined
- Nov 25, 2015
- Messages
- 244
- Reaction score
- 168
Hello,
We're using OpenAI for Voicemail and Call Transcription.
However, since quite some time (we don't often read back the transcriptions, so it took some time to notice this...) the transcriptions are failing on the GPT-4o-Transcribe-models.
The OpenAI Whisper model is still working properly.
In the event log, I get this:
Transcription stopped working: HTTP 403 (invalid_request_error: model_not_found) Project `proj_AoDY**` does not have access to model `gpt-4o-transcribe`
It seems the real error is 'model not found'.
I haven't changed anything in the OpenAI Platform-API (it's not even possible anymore to indicate which model may or may not be used).
So it seems that maybe the gpt-4o-transcribe model is no longer current?
Also, when changing to Whisper, I do get transcriptions but I no longer get summaries and/or sentiment.
Is anyone else seeing this as well?
We're using OpenAI for Voicemail and Call Transcription.
However, since quite some time (we don't often read back the transcriptions, so it took some time to notice this...) the transcriptions are failing on the GPT-4o-Transcribe-models.
The OpenAI Whisper model is still working properly.
In the event log, I get this:
Transcription stopped working: HTTP 403 (invalid_request_error: model_not_found) Project `proj_AoDY**` does not have access to model `gpt-4o-transcribe`
It seems the real error is 'model not found'.
I haven't changed anything in the OpenAI Platform-API (it's not even possible anymore to indicate which model may or may not be used).
So it seems that maybe the gpt-4o-transcribe model is no longer current?
Also, when changing to Whisper, I do get transcriptions but I no longer get summaries and/or sentiment.
Is anyone else seeing this as well?