--- name: azure-ai-transcription-py description: Azure AI Transcription SDK for Python. Use for audio-file transcription with timestamps; distinguish this client from streaming and asynchronous batch Speech APIs. risk: critical source: community date_added: '2026-02-27' --- # Azure AI Speech Transcription for Python ## When to Use - Transcribe an authorized audio file with the `azure-ai-transcription` client. - Review timestamped output and options supported by the installed SDK version. - For streaming or asynchronous batch jobs, select the appropriate Speech API instead. ## Prerequisites Use a project-pinned SDK in an isolated Python environment. Confirm the Speech resource, endpoint, permissions, audio consent, cost and retention before upload. Do not log keys, audio contents, or full transcripts by default. ## File Example ```python import os from azure.core.credentials import AzureKeyCredential from azure.ai.transcription import TranscriptionClient from azure.ai.transcription.models import TranscriptionContent, TranscriptionOptions client = TranscriptionClient( endpoint=os.environ["TRANSCRIPTION_ENDPOINT"], credential=AzureKeyCredential(os.environ["TRANSCRIPTION_KEY"]), ) with client: with open("audio.wav", "rb") as audio: result = client.transcribe(TranscriptionContent( definition=TranscriptionOptions(locales=["en-US"]), audio=audio, )) # Inspect result locally; handle empty output and service errors explicitly. ``` ## Authentication and API Boundary `DefaultAzureCredential` is also supported when the identity has the required resource role. The previous `begin_transcription` and `begin_stream_transcription` examples did not match this client. Do not invent a streaming adapter. ## Example Request “Transcribe this consented local recording; keep the transcript private.” ## Limitations Verify installed signatures and service limits against [Microsoft's SDK guide](https://learn.microsoft.com/en-us/python/api/overview/azure/ai-transcription-readme?view=azure-python). This illustration has not contacted Azure; availability, accuracy and costs require verification in the authorized environment.