2,372 questions with Azure Speech in Foundry Tools tags

Sort by: Updated
1 answer

MAI Transcribe 1.5: documented default "readability-optimized transcript" not observed — output is always verbatim, and transcribeStyle "readability" is rejected (400)

The fast transcription docs for MAI-Transcribe in LLM Speech API (updated 2026-07-21) say: Optionally, for mai-transcribe-1.5, you can specify the style of the transcript output by using transcribeStyle. By default, the model returns a…

Azure Speech in Foundry Tools
asked 2026-08-10T18:35:51.0566667+00:00
masaki 0 Reputation points
answered 2026-08-26T20:12:45.4933333+00:00
DataGuru42 16 Reputation points
1 answer

Speech to text is not working

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…

Azure Speech in Foundry Tools
asked 2026-08-18T07:26:17.07+00:00
commented 2026-08-26T20:06:25.4533333+00:00
kiro_nox 0 Reputation points
0 answers

Join: Microsoft Azure Q&A Champions Program

The Microsoft Azure Q&A Champions Program recognizes and empowers a global community of internal and external Azure experts who help customers succeed by providing high-quality, trusted answers on Microsoft Q&A. The program scales expert led…

Azure Speech in Foundry Tools
asked 2026-05-25T21:08:24.28+00:00
Ashley Swearingen 0 Reputation points Microsoft External Staff Moderator
edited the question 2026-08-26T18:46:32.8933333+00:00
Ashley Swearingen 0 Reputation points Microsoft External Staff Moderator
2 answers

Is audio-based training or acoustic adaptation available for Norwegian in Azure Custom Speech?

We are currently working on a real-time speech transcription and translation application using Azure Speech, and we are investigating whether it is possible to improve speech recognition accuracy for Norwegian by training or adapting a Custom Speech…

Azure Speech in Foundry Tools
asked 2026-08-21T18:39:45.49+00:00
Ignacio Massari 0 Reputation points
answered 2026-08-21T20:03:55.05+00:00
Jose Benjamin Solis Nolasco 11,621 Reputation points Volunteer Moderator
1 answer

ConversationTranscriber compatibility with Custom Speech

Does ConversationTranscriber in the JavaScript SDK ( microsoft-cognitiveservices-speech-sdk *v1.50.0) respect SpeechConfig.endpointId for custom speech models? The documentation lists endpoint_id as a property, but I want to confirm it works the same as…

Azure Speech in Foundry Tools
asked 2026-08-12T14:37:04.31+00:00
Tabarcea, Georgiana 0 Reputation points
answered 2026-08-14T01:50:53.47+00:00
Jerald Felix 18,680 Reputation points Volunteer Moderator
2 answers

what is the secure way to use Azure voice live service(from foundry) from browser

My concern is about the token being visible in browser . We tried relaying the service but avatar features are not that goood from pythin sdk

Azure Speech in Foundry Tools
asked 2026-08-13T21:03:59.2+00:00
Prasanna Sai Kommineni 90 Reputation points
answered 2026-08-13T21:06:21.7466667+00:00
Prasanna Sai Kommineni 90 Reputation points
0 answers

Fast Transcription returns English words for short Polish speech under forced pl-PL (regression since Aug 3, 2026)

We use the Azure Speech Fast Transcription REST API with a single, forced locale "locales": ["pl-PL"]: POST https://westeurope.api.cognitive.microsoft.com/speechtotext/transcriptions:transcribe?api-version=2024-11-15 Since August 3,…

Azure Speech in Foundry Tools
asked 2026-08-12T09:02:24.6933333+00:00
przemek 0 Reputation points
edited the question 2026-08-12T09:04:28.6366667+00:00
przemek 0 Reputation points
1 answer One of the answers was accepted by the question author.

zh-TW sapi phoneme: docs show pinyin examples, but the API returns an empty HTTP 400

The zh-TW phonetic-set tables in SSML phonetic alphabets have a column headed sapi Example (Bopomofo, Pinyin), and every row shows both notations — for example 知 (ㄓ, zhi 1) and 八 (ㄅㄚ or ㄅㄚˉ, ba 1). Reading that column as a list of valid ph values, I used…

Azure Speech in Foundry Tools
asked 2026-08-04T11:24:19.2633333+00:00
Mark C 20 Reputation points
accepted 2026-08-10T05:51:00.7933333+00:00
Mark C 20 Reputation points
2 answers

Configure Genesys TTS connector

Hello, We are trying to setup the genesys TTS connector to Microsoft azure TTS. https://help.genesys.cloud/articles/install-the-genesys-tts-connector-integration/#series-heading …

Azure Speech in Foundry Tools
asked 2026-07-14T12:58:37.3366667+00:00
Rob van der Poel (RF) 0 Reputation points
commented 2026-08-06T16:31:12.76+00:00
Karnam Venkata Rajeswari 5,250 Reputation points Microsoft External Staff Moderator
1 answer

When will Azure Pronunciation Assessment for en-GB leave preview status?

Hi, we're trying to build an English pronunciation practice app for German secondary schools using Azure Speech Pronunciation Assessment. We currently use en-US because en-GB is still in preview and returns empty Phonemes[].Phoneme labels (confirmed via…

Azure Speech in Foundry Tools
asked 2026-08-04T13:12:31.9166667+00:00
Frauke Leiss 0 Reputation points
answered 2026-08-06T06:19:32.25+00:00
Sridevi Machavarapu 33,740 Reputation points Microsoft External Staff Moderator
2 answers

Pronunciation Assessment header ignored — PronunciationAssessment block missing from NBest[0] despite correct configuration

iI am using the Speech to Text REST API for short audio with the Pronunciation Assessment feature enabled via the "Pronunciation-Assessment" header. Base speech recognition succeeds correctly (transcription, confidence score, and…

Azure Speech in Foundry Tools
asked 2026-07-12T19:30:15.7633333+00:00
Walter Roman 0 Reputation points
edited an answer 2026-08-04T16:40:20.7866667+00:00
Thanmayi Godithi 11,810 Reputation points Microsoft External Staff Moderator
1 answer

Request: Belgian Dutch (nl-BE) support for MAI-Voice-2 and MAI-Voice-2-Flash

Azure Speech already recognises Belgian Dutch as a separate locale and currently provides "nl-BE-DenaNeural" and "nl-BE-ArnaudNeural". However, the new MAI generation is currently available for Dutch through "nl-NL"…

Azure Speech in Foundry Tools
asked 2026-08-04T09:13:59.8933333+00:00
Dirk Delbaere 1 Reputation point
answered 2026-08-04T16:21:37.8+00:00
SRILAKSHMI C 19,735 Reputation points Microsoft External Staff Moderator
3 answers

Azure Communication Service's capability for audio injection to Teams Calls

We have designed a solution that successfully injects video, along with TTS-generated content, from an application into a Microsoft Teams call using Azure Communication Services (ACS). However, we have not been able to retrieve or inject audio back into…

Azure Speech in Foundry Tools
asked 2026-07-09T17:39:57.55+00:00
Trinanjan Majumder 0 Reputation points Microsoft Employee
commented 2026-08-03T03:24:34.3833333+00:00
Anshika Varshney 15,625 Reputation points Microsoft External Staff Moderator
1 answer

Azure Speech Fast Transcription (2025-10-15) returns masked word (*******) even with profanityFilterMode=None

Description We are using Azure AI Speech Fast Transcription REST API. Endpoint POST /speechtotext/transcriptions:transcribe?api-version=2025-10-15 Request { "audioUrl": "<audio-url>", "locales": [ …

Azure Speech in Foundry Tools
asked 2026-07-30T06:11:34.8933333+00:00
aassshhh 0 Reputation points
answered 2026-07-31T01:58:59.2166667+00:00
Jerald Felix 18,680 Reputation points Volunteer Moderator
1 answer

https://github.com/Azure-Samples/Cognitive-Speech-TTS/blob/master/Blog-Samples/Introducing-Dragon-HD-Omni/dragonhdomni_voice_list.json is broken

There are multiple issues. The json file is broken. All the AI generated voices were removed. Seriously? You removed half of the Omni voices without any announcement first? I'm seriously considering moving my backbone TTS service from Azure to other…

Azure Speech in Foundry Tools
asked 2026-07-21T03:29:35.7833333+00:00
Ming-Li Lin 20 Reputation points
commented 2026-07-29T04:32:26.5433333+00:00
Vinodh247-1375 43,996 Reputation points Volunteer Moderator
2 answers One of the answers was accepted by the question author.

EU AI Act Art. 50(2) — machine-readable marking of Azure Speech TTS output

Hello, We use Azure AI Speech (Neural TTS, region westeurope) in a language-learning product: (a) batch synthesis of static learning content at build time, and (b) real-time synthesis of dialogue replies in a live voice-conversation feature. Under the AI…

Azure Speech in Foundry Tools
asked 2026-07-26T17:16:19.9133333+00:00
Nikita 20 Reputation points
accepted 2026-07-28T07:28:49.1266667+00:00
Nikita 20 Reputation points
2 answers One of the answers was accepted by the question author.

Create more resources to do more TTS requests

I want to be able to do more TTS requests. Does creating a second resource for that make sense? We run critical for our business (phone calls) TTS pipeline and we absolutely have no room for failure.

Azure Speech in Foundry Tools
asked 2026-07-07T06:02:34.9666667+00:00
Siarhei Lukashenka 20 Reputation points
accepted 2026-07-27T04:17:59.6733333+00:00
Siarhei Lukashenka 20 Reputation points
1 answer

mslearn-ai-language 05-speech_tool

any one else have problems getting speech-agent to generate speech either aloud or to blob. I was able at one point to coax it to produce a temp link which I posted in another browser to recite hamlet. never wrote to file. I couldnt get past this to test…

Azure Speech in Foundry Tools
asked 2026-07-26T17:25:33.19+00:00
Leigh Walzer 0 Reputation points
edited an answer 2026-07-26T18:11:52.9033333+00:00
Kim Lehi Florendo 175 Reputation points
1 answer

How to fix azure.core.exceptions.ServiceResponseError: ('Connection aborted.', TimeoutError('The write operation timed out'))

I'm using this code for text to speech with python sdk But i from yesterday, i get 'Connection aborted.', TimeoutError'. I'm in django and i'm using azure as backend storage . This is my code: def ssml_to_audio( ssml_text, username): audio ,…

Azure Speech in Foundry Tools
asked 2024-03-13T02:15:40.55+00:00
Kokou AI 0 Reputation points
edited the question 2026-07-23T08:54:01.5633333+00:00
Caozhi Li 7,525 Reputation points Volunteer Moderator
3 answers

sl-SI - Petra and Rok - Incorrect lexical stress and abbreviation handling

Izdelek: Azure AI Speech - slovensko sl-SI - Petra, Rok Vsi glasovi napačno izgovarjajo pogoste besede zaradi napačnega leksikalnega naglasa. Primeri: gasilci (ga-SIL-ci namesto GA-sil-ci), pot, dom, les, požarna. Ročno moramo zapisati gasil-ci, po-ti,…

Azure Speech in Foundry Tools
asked 2026-07-20T12:01:07.52+00:00
Radio Sloven'c 0 Reputation points
answered 2026-07-21T13:58:32.9866667+00:00
Radio Sloven'c 0 Reputation points