feat(inputTranscription): add support for gpt-4o-transcribe & gpt-4o-mini-transcribe models, plus language & prompt parameters

[HVbajoria]

Purpose

• Added support for the new transcription models: gpt-4o-transcribe and gpt-4o-mini-transcribe
• Introduced optional language (ISO-639-1 format) and prompt parameters to the InputAudioTranscription interface
• Applied changes in both JavaScript and Python SDKs
• Updated the core library files and demonstrated usage in client_test under rtclient

Does this introduce a breaking change?

[ ] Yes
[x] No

Pull Request Type

What kind of change does this Pull Request introduce?

[ ] Bugfix
[x] Feature
[ ] Code style update (formatting, local variables)
[ ] Refactoring (no functional changes, no api changes)
[x] Documentation content changes
[ ] Other... Please describe:

How to Test

For JavaScript

• Get the code

  git clone https://github.com/Azure-Samples/aoai-realtime-audio-sdk.git
  cd aoai-realtime-audio-sdk
  cd javascript
  git checkout Added_Model_Support
  npm install

• Run tests

• Manually test usage with the updated inputTranscription settings under the standalone/test/client.spec.ts line 120:

  input_audio_transcription: {
           model:"gpt-4o-transcribe",
           language="en",
           prompt="expect words related to technology"
          },
  };

For Python

• Get the code

  git clone https://github.com/Azure-Samples/aoai-realtime-audio-sdk.git
  cd aoai-realtime-audio-sdk
  cd python
  cd samples
  git checkout Added_Model_Support
  pip3 install -r requirements.txt
  cd ..

• Test using client_test under rtclient:

  • Update the test snippet:

    input_audio_transcription = InputAudioTranscription(
        model="gpt-4o-transcribe",
        language="en",
        prompt="expect words related to technology"
    )

  • Try with different model values:

    • "whisper-1"
    • "gpt-4o-mini-transcribe"
    • "gpt-4o-transcribe"

What to Check

Verify that the following are valid:

• InputAudioTranscription in both Python and JavaScript accepts model, language, and prompt

• All new values (gpt-4o-transcribe, gpt-4o-mini-transcribe) are allowed and passed correctly

• Prompt format is respected based on the model:

  • Whisper: comma-separated keywords
  • GPT-4o models: free-text

• No regressions in existing behavior when using only whisper-1

• Functional parity between Python and JavaScript implementations

Other Information

• The updates are backward compatible
• Comments and type hints added where applicable
• Example test cases show use with different model values and prompt formats
• Enables developers to take advantage of OpenAI’s latest audio transcription capabilities across SDKs

[HVbajoria]

Hi @glecaros , @jpalvarezl, @trrwilson,

I have referred this document: https://learn.microsoft.com/en-us/azure/ai-services/openai/realtime-audio-reference

Under: RealtimeAudioInputTranscriptionSettings

Could you please check once?

[juliannicolas90]

This would be great to have!