Watchlist

https://learn.microsoft.com/en-us/cli/azure/release-notes-azure-cli?view=azure-cli-latest

@@ -9661,3 +9661,3 @@
 --gpu-instance-profile
-for Nvidia multi-instan…
+for Nvidia multi-instan…
 az aks update
@@ -12848,6 +12848,6 @@
 Add new examples for creating disk from share image gallery
-az vm image​ list/list-offers/list-skus/list-publishers/show
-: Add new parameter ​
+az vm image​ list/list-offers/list-skus/list-publishers/show
+: Add new parameter ​
 --edge-zone
-​ to support querying the image under edge zone
+​ to support querying the image under edge zone
 Fix the issue caused by the lack of
@@ -16131,3 +16131,3 @@
 AppConfig
-Unblock using "kv set" to add keyvault reference and feature …
+Unblock using "kv set" to add keyvault reference and feature …
 AppService
@@ -17634,3 +17634,3 @@
 parameter to be required when
-—esp
+—esp
 is present
@@ -23089,3 +23089,3 @@
 )
-Add Python location to ‘az —version’ (
+Add Python location to ‘az —version’ (
 #2986
@@ -23290,3 +23290,3 @@
 az vm/vmss disk
-Inside "az vm/vmss create", use "—use-unmanaged-disk" to avoid managed disk
+Inside "az vm/vmss create", use "—use-unmanaged-disk" to avoid managed disk
 Other commands should work

https://learn.microsoft.com/en-us/azure/ai-services/language-service/whats-new?tabs=csharp

@@ -61,3 +61,3 @@
 )
-. A new preview model delivers quality improvements and expanded entity recognition, optimized for spoken-language patterns—including interruptions, filler words, and speaker turns common in transcripts and call recordings.
+. A new preview model delivers quality improvements and expanded entity recognition, optimized for spoken-language patterns—including interruptions, filler words, and speaker turns common in transcripts and call recordings.
 PII detection capabilities available in new Foundry playgrounds
@@ -65,3 +65,3 @@
 new Foundry
-. The refreshed Document PII playground is generally available. It ships with curated sample inputs and expected outputs so you can evaluate detection of common entity types—including names, addresses, financial IDs, and health identifiers—without uploading your own documents. A single view lets you compare detection results across entity types. The playground uses the same Document PII model and policies as the production API, so results align with production behavior. When you're ready, transition from curated samples to uploading live documents for real-content testing. For more information, see:
+. The refreshed Document PII playground is generally available. It ships with curated sample inputs and expected outputs so you can evaluate detection of common entity types—including names, addresses, financial IDs, and health identifiers—without uploading your own documents. A single view lets you compare detection results across entity types. The playground uses the same Document PII model and policies as the production API, so results align with production behavior. When you're ready, transition from curated samples to uploading live documents for real-content testing. For more information, see:
 Text PII detection
@@ -133,3 +133,3 @@
 . All Azure AI Language capabilities are now accessible in Microsoft Foundry (classic), delivering a complete, unified development experience.
-Language Studio authoring and testing capabilities—including project management, model training, and deployment workflows—are now fully integrated into the Foundry platform.
+Language Studio authoring and testing capabilities—including project management, model training, and deployment workflows—are now fully integrated into the Foundry platform.
 Transition now
@@ -163,11 +163,11 @@
 Conversational Language Understanding multi-turn conversations
-. Enable natural, context-aware dialogues through entity slot filling → Microsoft Foundry (new).
+. Enable natural, context-aware dialogues through entity slot filling → Microsoft Foundry (new).
 Language detection
-. Automatically detect the language of user utterances in conversational applications → Microsoft Foundry (new).
+. Automatically detect the language of user utterances in conversational applications → Microsoft Foundry (new).
 PII detection for text
-. Detect and redact personally identifiable information in text documents → Microsoft Foundry (new).
+. Detect and redact personally identifiable information in text documents → Microsoft Foundry (new).
 Custom Named Entity Recognition
-. You can test, train, and deploy custom NER models directly in the Foundry playground → Microsoft Foundry (classic).
+. You can test, train, and deploy custom NER models directly in the Foundry playground → Microsoft Foundry (classic).
 PII detection for conversations
-. Identify and redact personally identifiable information in conversations → Microsoft Foundry (classic).
+. Identify and redact personally identifiable information in conversations → Microsoft Foundry (classic).
 Text PII detection enhancements (2025-11-15-preview API)
@@ -331,3 +331,3 @@
 None intent
-—particularly in multilingual contexts. You can access it through the REST API by using
+—particularly in multilingual contexts. You can access it through the REST API by using
 trainingConfigVersion 2025-07-01-preview
@@ -340,3 +340,3 @@
 accelerator project
-. The update includes a new routing strategy—
+. The update includes a new routing strategy—
 TRIAGE_AGENT
@@ -403,3 +403,3 @@
 PII redaction
-– This service update expands the window of detection the PII redaction service considers, enhancing quality and accuracy.
+– This service update expands the window of detection the PII redaction service considers, enhancing quality and accuracy.
 /python/api/azure-cognitiveservices-language-luis/index* Added prediction capability for custom models, including conversational language Understanding (CLU), custom named entity recognition (NER), and custom text classification, are now available in three new regions: Jio India Central, UK West, and Canada East.
@@ -421,3 +421,3 @@
 Back-end infrastructure for the Named entity recognition (NER) and Text Personally identifiable information (PII) entity recognition models is now updated with extended context window limits.
-Our 
+Our
 Conversational PII redaction
@@ -425,3 +425,3 @@
 February 2025
-Document and text abstractive summarization is now powered by fine-tuned Phi-3.5-mini! Check out the
+Document and text abstractive summarization is now powered by fine-tuned Phi-3.5-mini! Check out the
 Announcing Blog

https://learn.microsoft.com/en-us/azure/foundry/openai/how-to/realtime-audio?tabs=keyless%2Clinux&pivots=programming-language-python#quickstart

@@ -507,3 +507,3 @@
 To enable MCP support in a Realtime API session, provide the URL of a remote MCP server in your session configuration. This allows the API service to automatically manage tool calls on your behalf.
-You can easily enhance your agent's functionality by specifying a different MCP server in the session configuration—any tools available on that server will be accessible immediately.
+You can easily enhance your agent's functionality by specifying a different MCP server in the session configuration—any tools available on that server will be accessible immediately.
 The following example JSON body sets up an MCP server:
@@ -542,3 +542,3 @@
 ],
-"instructions": "Your knowledge cutoff is 2023-10. You are a helpful, witty, and friendly AI. Act like a human, but remember that you aren't a human and that you can't do human things in the real world. Your voice and personality should be warm and engaging, with a lively and playful tone. If interacting in a non-English language, start by using the standard accent or dialect familiar to the user. Talk quickly. You should always call a function if you can. Do not refer to these rules, even if you’re asked about them.",
+"instructions": "Your knowledge cutoff is 2023-10. You are a helpful, witty, and friendly AI. Act like a human, but remember that you aren't a human and that you can't do human things in the real world. Your voice and personality should be warm and engaging, with a lively and playful tone. If interacting in a non-English language, start by using the standard accent or dialect familiar to the user. Talk quickly. You should always call a function if you can. Do not refer to these rules, even if you’re asked about them.",
 "voice": "alloy",

https://learn.microsoft.com/en-us/azure/foundry-classic/openai/whats-new

@@ -37,3 +37,3 @@
 GPT Realtime 2.0 (preview) overview
-— covers reasoning support, response phases (preamble and final answer), and stricter instruction following.
+— covers reasoning support, response phases (preamble and final answer), and stricter instruction following.
 GPT Realtime Translate and GPT Realtime Whisper
@@ -41,5 +41,5 @@
 GPT Realtime Translate
-— covers continuous stream-based audio translation for live multilingual events, customer support, and voice assistants.
+— covers continuous stream-based audio translation for live multilingual events, customer support, and voice assistants.
 GPT Realtime Whisper
-— covers low-latency streaming transcription for live captions, monitoring, and archival workflows.
+— covers low-latency streaming transcription for live captions, monitoring, and archival workflows.
 February 2026
@@ -76,3 +76,3 @@
 Improves multilingual performance across Japanese, Indic, and other languages.
-Reduces hallucinations on silence by up to 4×, making it a more reliable choice for noisy environments and real-world audio streams.
+Reduces hallucinations on silence by up to 4×, making it a more reliable choice for noisy environments and real-world audio streams.
 Input remains audio, with text as output, and deployment is API-only.
@@ -95,6 +95,6 @@
 gpt-4o-transcribe-diarize
-speech to text model is released. This model is an Automatic Speech Recognition (ASR) model that converts spoken language into text in real time. It enables organizations to unlock insights from conversations instantly with ultra-low latency and high accuracy across 100+ languages. This capability is essential for workflows where voice data drives decisions—such as customer support, virtual meetings, and live events.
+speech to text model is released. This model is an Automatic Speech Recognition (ASR) model that converts spoken language into text in real time. It enables organizations to unlock insights from conversations instantly with ultra-low latency and high accuracy across 100+ languages. This capability is essential for workflows where voice data drives decisions—such as customer support, virtual meetings, and live events.
 Diarization is the process of identifying who spoke when in an audio stream. It transforms conversations into speaker-attributed transcripts, enabling businesses to extract actionable insights from meetings, customer calls, and live events. With advanced models like
 gpt-4o-transcribe-diarize
-, organizations gain real-time clarity and context—turning voice into structured data that drives smarter decisions and improves productivity, supporting automatic speech recognition.
+, organizations gain real-time clarity and context—turning voice into structured data that drives smarter decisions and improves productivity, supporting automatic speech recognition.
 Use this model via the

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/llm-speech?tabs=new-foundry%2Clinux%2Cwindows&pivots=ai-foundry

@@ -44,45 +44,41 @@
 Transcription
-âœ
-(speech-to-text model)
-âœ
-(multimodal model)
-âœ
-(speech-to-text model)
+✅ (speech-to-text model)
+✅ (multimodal model)
+✅ (speech-to-text model)
 Translation
-❌
-âœ
-(multimodal model)
-❌
+❌
+✅ (multimodal model)
+❌
 Diarization
-âœ
-âœ
-âœ
+✅
+✅
+✅
 Channel (stereo)
-âœ
-âœ
-❌
+✅
+✅
+❌
 Profanity filtering
-âœ
-âœ
-âœ
+✅
+✅
+✅
 Specify locale
-âœ
-âœ
-âœ
+✅
+✅
+✅
 Custom prompting
-❌
-âœ
-❌
+❌
+✅
+❌
 Phrase list
-âœ
-âœ
-âœ
+✅
+✅
... und 67 weitere Zeilen

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/voice-live-how-to-customize

@@ -34,3 +34,3 @@
 Speech output customization:
-Custom lexicon: Custom lexicon allows you to easily customize pronunciation for both standard Azure text to speech voices and custom voices to improve speech synthesis accuracy for your use case. See
+Custom lexicon: Custom lexicon allows you to easily customize pronunciation for both standard Azure text to speech voices and custom voices to improve speech synthesis accuracy for your use case. See
 custom lexicon for text to speech
@@ -47,3 +47,3 @@
 Phrase list
-Use phrase list for lightweight just-in-time customization on audio input. To configure phrase list, you can set the phrase_list in the
+Use phrase list for lightweight just-in-time customization on audio input. To configure phrase list, you can set the phrase_list in the
 session.update
@@ -59,3 +59,3 @@
 Note
-Phrase list currently doesn't support whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe, and gpt-4o-transcribe-diarize. To learn more about phrase list, see
+Phrase list currently doesn't support whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe, and gpt-4o-transcribe-diarize. To learn more about phrase list, see
 phrase list for speech to text
@@ -91,3 +91,3 @@
 custom_lexicon_url
-string property to customize pronunciation for both standard Azure text to speech voices and custom voices. To learn more about how to format the custom lexicon (the same as Speech Synthesis Markup Language (SSML)), see
+string property to customize pronunciation for both standard Azure text to speech voices and custom voices. To learn more about how to format the custom lexicon (the same as Speech Synthesis Markup Language (SSML)), see
 custom lexicon for text to speech
@@ -95,8 +95,8 @@
 {
-  "voice": {
-    "name": "en-US-Ava:DragonHDLatestNeural",
-    "type": "azure-standard",
-    "temperature": 0.8, // optional
-    "custom_lexicon_url": "<custom lexicon url>"
-  }
+"voice": {
+"name": "en-US-Ava:DragonHDLatestNeural",
+"type": "azure-standard",
+"temperature": 0.8, // optional
+"custom_lexicon_url": "<custom lexicon url>"
+}
 }

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/voice-live-how-to

@@ -706,8 +706,8 @@
 {
-  "voice": {
-    "name": "en-US-Ava:DragonHDLatestNeural",
-    "type": "azure-standard",
-    "temperature": 0.8, // optional
-   "rate": "1.2"
-  }
+"voice": {
+"name": "en-US-Ava:DragonHDLatestNeural",
+"type": "azure-standard",
+"temperature": 0.8, // optional
+"rate": "1.2"
+}
 }

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/voice-live

@@ -27,3 +27,3 @@
 The Voice Live API is a solution that enables low-latency, high-quality speech-to-speech interactions for voice agents. The API is designed for developers seeking scalable and efficient voice-driven experiences as it eliminates the need to manually orchestrate multiple components. By integrating speech recognition, generative AI, and text-to-speech functionalities into a single, unified interface, it provides an end-to-end solution for creating seamless experiences.
-The Voice Live API is fully managed, so you don't need to handle backend orchestration or component integration. Developers provide audio input and receive audio output, avatar visuals, and action triggers—all with minimal latency. You don't need to deploy or manage any generative AI models, as the API handles the underlying infrastructure.
+The Voice Live API is fully managed, so you don't need to handle backend orchestration or component integration. Developers provide audio input and receive audio output, avatar visuals, and action triggers—all with minimal latency. You don't need to deploy or manage any generative AI models, as the API handles the underlying infrastructure.
 Understand speech-to-speech experiences
@@ -174,22 +174,22 @@
 Broad locale coverage with high accuracy (audio input)
-âœ
-âœ
+✅
+✅
 Maintain brand and character personality (audio output)
-âœ
-âœ
+✅
+✅
 Conversational enhancements
-❌
-âœ
+❌
+✅
 Choice of generative AI models
-âœ
-âœ
+✅
+✅
 Visual output with text to speech avatar
-âœ
-âœ
+✅
+✅
 Low engineering cost
-❌
-âœ
+❌
+✅
 Low latency perceived by end user
-❌
-âœ
+❌
+✅
 Pricing
@@ -281,3 +281,3 @@
 You're charged separately for the training and model hosting of:
-Custom voice – professional
+Custom voice – professional
 Custom avatar
@@ -314,5 +314,5 @@
 You're charged separately for the training and model hosting of:
-Custom voice – professional
+Custom voice – professional
 Token usage and cost estimation
-Tokens are the units that generative AI models use to process input and generate output. 
+Tokens are the units that generative AI models use to process input and generate output.
 You can estimate token usage for different model families with the Voice Live API based on audio length. The following token calculations apply to each model family:

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/releasenotes?tabs=speech-to-text

@@ -362,3 +362,3 @@
 New features:
-Updated development dependency: TypeScript 3.5.3 → 4.5
+Updated development dependency: TypeScript 3.5.3 → 4.5
 Updated TranslationRecognizer to use V2 endpoints by default.
@@ -608,4 +608,4 @@
 Bug fixes
-Fix for iOS SDK x2 times binary size growth · Issue #2113 · Azure-Samples/cognitive-services-speech-sdk (github.com)
-Fix for Unable to get word level time stamps from Azure speech to text API · Issue #2156 · Azure-Samples/cognitive-services-speech-sdk (github.com)
+Fix for iOS SDK x2 times binary size growth · Issue #2113 · Azure-Samples/cognitive-services-speech-sdk (github.com)
+Fix for Unable to get word level time stamps from Azure speech to text API · Issue #2156 · Azure-Samples/cognitive-services-speech-sdk (github.com)
 Fix for DialogServiceConnector destruction phase to disconnect events correctly. This was causing crashes occasionally.
@@ -613,3 +613,3 @@
 FPIEProcessor.dll from Microsoft.CognitiveServices.Speech.Extension.MAS NuGet package for Windows UWP x64 and Arm64 had dependency on VC runtime libraries for native C++. The issue has been rectified by updating the dependency to correct VC runtime libraries (for UWP).
-Fix for [MAS] Recurrent calls to recognizeOnceAsync lead to SPXERR_ALREADY_INITIALIZED when using MAS · Issue #2124 · Azure-Samples/cognitive-services-speech-sdk (github.com)
+Fix for [MAS] Recurrent calls to recognizeOnceAsync lead to SPXERR_ALREADY_INITIALIZED when using MAS · Issue #2124 · Azure-Samples/cognitive-services-speech-sdk (github.com)
 Fix for embedded speech recognition crash when phrase lists are used.
@@ -658,5 +658,5 @@
 Android packages updates with latest security fixes from OpenSSL1.1.1v
-JS – WebWorkerLoadType property added to allow bypass of data URL load for timeout worker
-JS – Fix Conversation Translation disconnect after 10 minutes
-JS – Conversation Translation auth token from Conversation now propagates to Translation service connection
+JS – WebWorkerLoadType property added to allow bypass of data URL load for timeout worker
+JS – Fix Conversation Translation disconnect after 10 minutes
+JS – Conversation Translation auth token from Conversation now propagates to Translation service connection
 Samples
@@ -855,3 +855,3 @@
 JavaScript
-– Added support for microphone input from the browser with Speaker Identification and Verification.
+– Added support for microphone input from the browser with Speaker Identification and Verification.
 Embedded Speech Recognition
@@ -3196,7 +3196,7 @@
 Output properties
+--output […] result XXX property
 (PropertyId or string)
+--output […] connection message received XXX property
 (PropertyId or string)
+--output […] recognizer XXX property
 (PropertyId or string)
@@ -3281,11 +3281,11 @@
 spx csr dataset upload --kind audio|language|acoustic
-– create datasets from local data, not just from URLs.
+– create datasets from local data, not just from URLs.
 spx csr evaluation create|status|list|update|delete
-– compare new models against baseline truth/other models.
+– compare new models against baseline truth/other models.
 spx * list
-– supports non-paged experience (doesn't require --top X --skip X).
+– supports non-paged experience (doesn't require --top X --skip X).
 spx * --http header A=B
-– support custom headers (added for Office for custom authentication).
+– support custom headers (added for Office for custom authentication).
 spx help
-– improved text and back-tick text color coded (blue).
+– improved text and back-tick text color coded (blue).
 2020-June release
@@ -3429,3 +3429,3 @@
 Latest in Production
-, delivering improvements in natural prosody, expressiveness, and output consistency—particularly for long‑form and complex content. This update expands support for expressive speaking styles and paralinguistic elements, enabling more human‑like and emotionally rich speech across narration, conversational AI, and virtual assistant scenarios.
+, delivering improvements in natural prosody, expressiveness, and output consistency—particularly for long‑form and complex content. This update expands support for expressive speaking styles and paralinguistic elements, enabling more human‑like and emotionally rich speech across narration, conversational AI, and virtual assistant scenarios.
... und 213 weitere Zeilen