[VoiceLive] Add LLMVoice and AzurePlatformVoice support, update endpoint to voice-live (#42997)

* [VoiceLive] Add async function-calling agent sample

* add phrase list

* fix typo

* Update sdk/ai/azure-ai-voicelive/samples/async_function_calling_sample.py

Co-authored-by: Copilot <[email protected]>

* Update sdk/ai/azure-ai-voicelive/samples/async_function_calling_sample.py

Co-authored-by: Copilot <[email protected]>

* update

* fix typo

* update changelog

* update

* remove breaking change section

* update changelog

* fix change log

* revert changelog I lost

* update version and change log

* enable type verification

* update

* update models

* fix pylint issue

---------

Co-authored-by: Xiting Zhang <[email protected]>
Co-authored-by: Copilot <[email protected]>

Author: 3 people

sdk/ai/azure-ai-voicelive/CHANGELOG.md

@@ -4,11 +4,20 @@
 
 ### Features Added
 
-- Phrase list
+- **Transcription improvement**: Added phrase list
+- **New Voice Types**: Added `AzurePlatformVoice` and `LLMVoice` classes
+- **Enhanced Speech Detection**: Added `AzureSemanticVadServer` class
+- **Improved Function Calling**: Enhanced async function calling sample with better error handling
 
 ### Breaking Changes
 
-- Removed `custom_model` and `enabled` from `AudioInputTranscriptionSettings`.
+- **Transcription**: Removed `custom_model` and `enabled` from `AudioInputTranscriptionSettings`.
+- **Async Authentication**: Fixed credential handling for async scenarios
+- **Model Serialization**: Improved error handling and deserialization
+
+### Other Changes
+
+- **Code Modernization**: Updated type annotations throughout
 
 ## 1.0.0b2 (2025-09-10)
 

sdk/ai/azure-ai-voicelive/apiview-properties.json

@@ -15,10 +15,12 @@
         "azure.ai.voicelive.models.TurnDetection": "VoiceLive.TurnDetection",
         "azure.ai.voicelive.models.AzureMultilingualSemanticVad": "VoiceLive.AzureMultilingualSemanticVad",
         "azure.ai.voicelive.models.AzurePersonalVoice": "VoiceLive.AzurePersonalVoice",
+        "azure.ai.voicelive.models.AzurePlatformVoice": "VoiceLive.AzurePlatformVoice",
         "azure.ai.voicelive.models.EOUDetection": "VoiceLive.EOUDetection",
         "azure.ai.voicelive.models.AzureSemanticDetection": "VoiceLive.AzureSemanticDetection",
         "azure.ai.voicelive.models.AzureSemanticDetectionMultilingual": "VoiceLive.AzureSemanticDetectionMultilingual",
         "azure.ai.voicelive.models.AzureSemanticVad": "VoiceLive.AzureSemanticVad",
+        "azure.ai.voicelive.models.AzureSemanticVadServer": "VoiceLive.AzureSemanticVadServer",
         "azure.ai.voicelive.models.AzureStandardVoice": "VoiceLive.AzureStandardVoice",
         "azure.ai.voicelive.models.ClientEvent": "VoiceLive.ClientEvent",
         "azure.ai.voicelive.models.ClientEventConversationItemCreate": "VoiceLive.ClientEventConversationItemCreate",
@@ -48,6 +50,7 @@
         "azure.ai.voicelive.models.InputAudioContentPart": "VoiceLive.InputAudioContentPart",
         "azure.ai.voicelive.models.InputTextContentPart": "VoiceLive.InputTextContentPart",
         "azure.ai.voicelive.models.InputTokenDetails": "VoiceLive.InputTokenDetails",
+        "azure.ai.voicelive.models.LLMVoice": "VoiceLive.LLMVoice",
         "azure.ai.voicelive.models.LogProbProperties": "VoiceLive.LogProbProperties",
         "azure.ai.voicelive.models.NoTurnDetection": "VoiceLive.NoTurnDetection",
         "azure.ai.voicelive.models.OpenAIVoice": "VoiceLive.OpenAIVoice",
@@ -91,7 +94,7 @@
         "azure.ai.voicelive.models.ServerEventResponseContentPartDone": "VoiceLive.ServerEventResponseContentPartDone",
         "azure.ai.voicelive.models.ServerEventResponseCreated": "VoiceLive.ServerEventResponseCreated",
         "azure.ai.voicelive.models.ServerEventResponseDone": "VoiceLive.ServerEventResponseDone",
-        "azure.ai.voicelive.models.ServerEventResponseFunctionCallArgumentsDelta": "VoiceLive.ServerEventResponseFunctionCallArgumentsDelta",
+                "azure.ai.voicelive.models.ServerEventResponseFunctionCallArgumentsDelta": "VoiceLive.ServerEventResponseFunctionCallArgumentsDelta",
         "azure.ai.voicelive.models.ServerEventResponseFunctionCallArgumentsDone": "VoiceLive.ServerEventResponseFunctionCallArgumentsDone",
         "azure.ai.voicelive.models.ServerEventResponseOutputItemAdded": "VoiceLive.ServerEventResponseOutputItemAdded",
         "azure.ai.voicelive.models.ServerEventResponseOutputItemDone": "VoiceLive.ServerEventResponseOutputItemDone",

sdk/ai/azure-ai-voicelive/azure/ai/voicelive/_patch.py

@@ -17,9 +17,7 @@
 from typing import (
     TYPE_CHECKING,
     Any,
-    Dict,
     Iterator,
-    List,
     Mapping,
     Optional,
     Sequence,
@@ -57,9 +55,11 @@
     try:
         from websockets.typing import Subprotocol as WSSubprotocol  # runtime if available
     except Exception:
+
         class WSSubprotocol(str):  # fallback, keeps runtime simple
             pass
 
+
 if TYPE_CHECKING:
     # Not imported at runtime; only for type checkers (mypy/pyright).
     from websockets.sync.client import ClientConnection as _WSClientConnection
@@ -69,7 +69,7 @@ class WSSubprotocol(str):  # fallback, keeps runtime simple
 else:
     from typing_extensions import NotRequired  # noqa: F401
 
-__all__: List[str] = [
+__all__: list[str] = [
     "connect",
     "WebsocketConnectionOptions",
     "ConnectionError",
@@ -339,7 +339,7 @@ def clear(self, *, event_id: Optional[str] = None) -> None:
         :return: None
         :rtype: None
         """
-        event: Dict[str, Any] = {"type": "output_audio_buffer.clear"}
+        event: dict[str, Any] = {"type": "output_audio_buffer.clear"}
         if event_id:
             event["event_id"] = event_id
         self._connection.send(event)
@@ -357,11 +357,7 @@ def __init__(self, connection: "VoiceLiveConnection") -> None:
         self._connection = connection
 
     def create(
-        self,
-        *,
-        item: Mapping[str, Any],
-        previous_item_id: Optional[str] = None,
-        event_id: Optional[str] = None
+        self, *, item: Mapping[str, Any], previous_item_id: Optional[str] = None, event_id: Optional[str] = None
     ) -> None:
         """Create a new conversation item.
 
@@ -477,7 +473,7 @@ def update(self, *, session: Mapping[str, Any], event_id: Optional[str] = None)
         :return: None
         :rtype: None
         """
-        event: Dict[str, Any] = {"type": "transcription_session.update", "session": dict(session)}
+        event: dict[str, Any] = {"type": "transcription_session.update", "session": dict(session)}
         if event_id:
             event["event_id"] = event_id
         self._connection.send(event)
@@ -674,14 +670,14 @@ def __enter__(self) -> VoiceLiveConnection:
 
             # Build headers as str->str and use list of tuples to satisfy HeadersLike
             extra_headers_map: Mapping[str, Any] = self.__extra_headers or {}
-            merged_headers: Dict[str, str] = {
+            merged_headers: dict[str, str] = {
                 **self._get_auth_headers(),
                 **{str(k): str(v) for k, v in extra_headers_map.items()},
             }
-            headers_like: List[Tuple[str, str]] = list(merged_headers.items())
+            headers_like: list[Tuple[str, str]] = list(merged_headers.items())
 
             # Build kwargs for websockets; avoid dict(Optional[...])
-            ws_kwargs: Dict[str, Any] = {}
+            ws_kwargs: dict[str, Any] = {}
             if self.__connection_options is not None:
                 ws_kwargs.update(cast(Mapping[str, Any], self.__connection_options))
 
@@ -715,7 +711,7 @@ def __exit__(self, exc_type, exc, exc_tb) -> None:
         if self.__connection is not None:
             self.__connection.close()
 
-    def _get_auth_headers(self) -> Dict[str, str]:
+    def _get_auth_headers(self) -> dict[str, str]:
         """Get authentication headers for WebSocket connection.
 
         :return: A dictionary containing authentication headers.
@@ -735,7 +731,7 @@ def _prepare_url(self) -> str:
         parsed = urlparse(self._endpoint)
         scheme = "wss" if parsed.scheme == "https" else ("ws" if parsed.scheme == "http" else parsed.scheme)
 
-        params: Dict[str, str] = {"model": self.__model, "api-version": self.__api_version}
+        params: dict[str, str] = {"model": self.__model, "api-version": self.__api_version}
         extra_query: Mapping[str, Any] = self.__extra_query or {}
         for k, v in extra_query.items():
             params[str(k)] = str(v)
@@ -746,7 +742,7 @@ def _prepare_url(self) -> str:
             if key not in params:
                 params[key] = value_list[0] if value_list else ""
 
-        path = parsed.path.rstrip("/") + "/voice-agent/realtime"
+        path = parsed.path.rstrip("/") + "/voice-live/realtime"
         return urlunparse((scheme, parsed.netloc, path, parsed.params, urlencode(params), parsed.fragment))
 
 

sdk/ai/azure-ai-voicelive/azure/ai/voicelive/_types.py

@@ -10,5 +10,7 @@
 
 if TYPE_CHECKING:
     from . import models as _models
-Voice = Union[str, "_models.OAIVoice", "_models.OpenAIVoice", "_models.AzureVoice", "_models.Phi4mmVoice"]
+Voice = Union[
+    str, "_models.OAIVoice", "_models.OpenAIVoice", "_models.AzureVoice", str, "_models.Phi4mmVoice", "_models.LLMVoice"
+]
 ToolChoice = Union[str, "_models.ToolChoiceLiteral", "_models.ToolChoiceObject"]