Problem with wrong audio type

[iEnki]

Hey all,

i can´t fix this Problem. ffmpeg installed finde, the audi file is saved as .wav but still this error stops my transcription. Any Ideas for this fix. Iam on windows system, maybe there are codec error?

Error: openai.error.InvalidRequestError: Invalid file format. Supported formats: ['m4a', 'mp3', 'webm', 'mp4', 'mpga', 'wav', 'mpeg']

[iwmo]

yeah...having same problem on MacOs

[Jayqianonedream]

Same to me, try both on win and mac

[yinyijie]

same error here.

[VikasSharma707]

Solved solution go to this repo https://github.com/VikasSharma707/chatgpt-api-whisper-api-voice-assistant

[Jupalaja]

Hi guys, this worked for me:

from pydub import AudioSegment

audio_file_wav = open(audio, "rb")
audio_file_mp3 = AudioSegment.from_wav(audio_file_wav).export("audio.mp3", format="mp3")
transcript = openai.Audio.transcribe("whisper-1", audio_file_mp3)

I'm using pydub to convert the audio file to wav, this export the mp3 file that can be send to the Whisper API without a problem

[yinyijie]

Hi guys, this worked for me:

from pydub import AudioSegment

audio_file_wav = open(audio, "rb")
audio_file_mp3 = AudioSegment.from_wav(audio_file_wav).export("audio.mp3", format="mp3")
transcript = openai.Audio.transcribe("whisper-1", audio_file_mp3)

I'm using pydub to convert the audio file to wav, this export the mp3 file that can be send to the Whisper API without a problem

Thanks, it resolved the issue.

[eervin123]

Was having this problem as well. I found this solution, similar concept to the one above. We are just adding an audio file extension.

https://github.com/gradio-app/gradio/issues/3479

def transcribe(audio):
    os.rename(audio, audio + '.wav')
    file = open(audio + '.wav', "rb")
    return openai.Audio.transcribe("whisper-1", file).text

[hackingthemarkets]

Thanks for the rename suggestion, added this to repo :). I'm guessing the API didn't initially validate the extension, but they added this after I recorded.