QuestionQ35

VerifiedPlan and manage an Azure AI solution

You have an Azure Speech in Foundry Tools resource that hosts a custom speech-to-text model deployed to a custom endpoint, and an agent uses that endpoint for real-time speech recognition.

The custom speech-to-text model is nearing its expiration date.

What is the expected behavior once the model expires?

Explanation

Custom Speech models and their deployed endpoints follow a defined lifecycle with an expiration date. When a custom model reaches that date, the associated custom endpoint does not shut down or begin returning errors; instead, the Speech service automatically falls back to using the most recent base model for the same locale, so recognition requests keep succeeding, only losing the custom adaptation rather than failing outright. The endpoint is not deleted automatically, and expired custom models are not retained indefinitely for continued use without manual intervention.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!