Speech to text is not working

Mahesh Pratap Singh alias Yogi M. P. Singh 0 Reputation points
2026-08-18T07:26:17.07+00:00

Subject: Azure Speech-to-Text (Speech Service) not working / recognition failing

Service affected: Azure AI Speech (Speech-to-Text / Speech SDK)

Issue summary: Speech-to-text transcription is not working as expected. [Choose the one that applies and delete the others:]

  • No transcription output is returned at all (empty/null result)
  • Requests fail with an error (see error details below)
  • Audio is submitted but recognition never completes / times out
  • Recognized text is inaccurate or garbled
  • Service was working previously and stopped working on [date]

Environment details:

  • Azure region: [e.g., East US, West Europe]
  • Speech resource name: [your resource name]
  • SDK / API used: [Speech SDK (language: C#/Python/JS/Java), REST API, or Speech Studio]
  • SDK version: [version number]
  • Platform: [Windows/Linux/macOS/mobile/web], OS version: [version]
  • Audio input source: [microphone, .wav file, streaming, phone call, etc.]
  • Audio format: [sample rate, channels, bit depth, e.g., 16kHz, mono, PCM WAV]
  • Language/locale configured: [e.g., en-US]Subject: Azure Speech-to-Text (Speech Service) not working / recognition failing Service affected: Azure AI Speech (Speech-to-Text / Speech SDK) Issue summary:
    Speech-to-text transcription is not working as expected. [Choose the one that applies and delete the others:]
    • No transcription output is returned at all (empty/null result)
    • Requests fail with an error (see error details below)
    • Audio is submitted but recognition never completes / times out
    • Recognized text is inaccurate or garbled
    • Service was working previously and stopped working on [date]
    Environment details:
    • Azure region: [e.g., East US, West Europe]
    • Speech resource name: [your resource name]
    • SDK / API used: [Speech SDK (language: C#/Python/JS/Java), REST API, or Speech Studio]
    • SDK version: [version number]
    • Platform: [Windows/Linux/macOS/mobile/web], OS version: [version]
    • Audio input source: [microphone, .wav file, streaming, phone call, etc.]
    • Audio format: [sample rate, channels, bit depth, e.g., 16kHz, mono, PCM WAV]
    • Language/locale configured: [e.g., en-US]
Azure Speech in Foundry Tools
0 comments

2 answers

Sort by: Most helpful
Most helpful Newest Oldest
  1. Thanmayi Godithi 11,825 Reputation points Microsoft External Staff Moderator
    2026-09-02T14:40:01.3266667+00:00

    Hi ˇMahesh Pratap Singh alias Yogi M. P. Singh ,

    Thank you for sharing the additional details and your observations.

    From your description, this appears to be different from an application-level Speech SDK issue. You mentioned that Windows Voice Typing (Win + H) is listening but never returns transcribed text, eventually timing out, and that you are seeing a service availability message when attempting related connectivity checks.

    However, the information provided in this thread is not sufficient to confirm a service-side outage. To help determine whether the issue is related to the client device, network connectivity, or a backend service problem, could you please provide:

    • Your Windows version and build.
    • Whether the issue occurs on multiple networks (for example, home network vs. mobile hotspot).
    • Whether other cloud-powered Windows features are working normally.
    • A screenshot of the exact "Our services aren't available right now" message you referenced.

    For developers using Azure Speech, a useful validation step is to run the official Speech-to-Text quickstart against the same Speech resource to determine whether the issue is specific to the application or affects the service more broadly.

    If recognition requests are being cancelled or timing out, enabling Speech SDK logging and capturing the cancellation details can provide additional diagnostic information.

    Thanks again for reporting what you're seeing. Additional details will help determine whether this is a local voice typing issue or a broader service connectivity problem.

    Was this answer helpful?

    0 comments
  2. Amanda Zhu 160 Reputation points
    2026-08-18T16:12:06.09+00:00

    You haven't filled out the entire troubleshooting template, so there isn’t enough information to identify a Speech service failure yet.

    To narrow down your issue, I would make 2 checks:

    • Run the same input through the documented speech-to-text quickstart using the same Speech resource, region, and locale. This separates an application/audio-pipeline issue from resource configuration.
    • If the result is Canceled or recognition stalls, capture the cancellation details and enable Speech SDK logging. Include the SDK version, platform, audio format, region, and the exact ErrorCode/ErrorDetails in the question.

    If this is long-running or streaming audio, also verify that the app is using continuous recognition rather than RecognizeOnceAsync, which is intended for a single utterance.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.