Inconsistency in the "Speech to text" documentation

Documentation
,
1

Hello,

I’m reporting a small inconsistency in the “Speech to text” documentation.

In the documentation, there is this excerpt:

from openai import OpenAI

client = OpenAI()
audio_file = open("/path/to/file/speech.mp3", "rb")

transcription = client.audio.transcriptions.create(
    model="gpt-4o-transcribe", 
    file=audio_file, 
    response_format="text"
)

print(transcription.text)

However, running this leads to an AttributeError, because transcription will actually be of type str. This is expected, as confirmed by the implementation of audio.transcriptions.create which expects a str return when response_format equals "text".

A quick fix would be to change the print(transcription.text) to print(transcription) in the documentation.

thx!