[VoiceLive] Add changes for 2026-06-01-preview

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

@@ -1,15 +1,29 @@
 # Release History
 
-## 1.1.0-beta.1 (Unreleased)
+## 1.1.0-beta.1 (2026-06-02)
 
 ### Features Added
 
-### Breaking Changes
-
-### Bugs Fixed
+- **WebRTC SDP negotiation** for browser/native WebRTC clients:
+  - New `ClientEventRtcCallSdpCreate` (`rtc.call.sdp.create`) to send an SDP offer with an optional initial `VoiceLiveSessionOptions`
+  - New `ServerEventRtcCallSdpCreated` (`rtc.call.sdp.created`) carrying the SDP answer and `rtcCallId`
+  - New `ServerEventRtcCallError` (`rtc.call.error`) with structured `RtcCallErrorDetails` (type/code/message) and originating operation
+- **Streaming input text** into a conversation item:
+  - New `ClientEventInputTextDelta` (`input_text.delta`) and `ClientEventInputTextDone` (`input_text.done`), each with optional `contentIndex`
+- **Output audio buffer playback lifecycle**:
+  - New `ServerEventOutputAudioBufferStarted` (`output_audio_buffer.started`) and `ServerEventOutputAudioBufferStopped` (`output_audio_buffer.stopped`), each exposing `responseId`
+- **Smart audio-based end-of-turn detection**: new `SmartEndOfTurnDetection` (`smart_end_of_turn_detection`) `EouDetection` variant with `thresholdLevel` and `timeoutMs`; new `EouDetectionModel.SMART_END_OF_TURN_DETECTION` value
+- **Echo cancellation reference source**: `AudioEchoCancellation` is now `@Fluent` and exposes `referenceSource` (new `EchoCancellationReferenceSource` enum with `SERVER` / `CLIENT`) and `channels` for stereo input where channel 1 is the client-supplied echo reference
+- **Azure realtime native voices**: new `AzureRealtimeNativeVoice` (type `azure-realtime-native`) and `AzureRealtimeNativeVoiceName` expandable enum (`AARTI`, `ANDREW`, `AVA`, `DENISE`, `DIYA`, `ELSA`, `FLORIAN`, `FRANCISCA`, `MEERA`, `XIAOXIAO`, `YUNXI`, `XIMENA`) for use with the `azure-realtime` model
+- **Parallel tool calls**: `VoiceLiveSessionOptions.setParallelToolCalls(Boolean)` / `isParallelToolCalls()` (and matching getter on `VoiceLiveSessionResponse`) to control whether the model may invoke tools in parallel
+- **Hosted agent invocation passthrough**:
+  - `ResponseCreateParams.setInvokeInput(Map<String, BinaryData>)` to attach input data for a hosted agent invocation on a single response request (preview)
+  - New `ServerEventResponseInvocationDelta` (`response.invocation.delta`) passes through non-speech SSE events from the hosted agent
 
 ### Other Changes
 
+- Regenerated against `2026-06-01-preview` VoiceLive TypeSpec. The default service API version used by `VoiceLiveClientBuilder` is unchanged (`V2026_04_10`, GA).
+
 ## 1.0.0 (2026-06-01)
 
 This is the first General Availability (GA) release of the Azure VoiceLive client library for Java.

sdk/voicelive/azure-ai-voicelive/cspell.json

@@ -0,0 +1,36 @@
+{
+  "version": "0.2",
+  "language": "en",
+  "words": [
+    "Aarti",
+    "AARTI",
+    "aiservices",
+    "AIservices",
+    "Dexec",
+    "Diya",
+    "DIYA",
+    "filler",
+    "FILLER",
+    "foundry",
+    "FOUNDRY",
+    "genai",
+    "GENAI",
+    "HDOMNI",
+    "Meera",
+    "MEERA",
+    "SSML",
+    "Unpooled",
+    "viseme",
+    "VISEME",
+    "webrtc",
+    "WEBRTC",
+    "Xiaoxiao",
+    "XIAOXIAO",
+    "Ximena",
+    "XIMENA",
+    "xhigh",
+    "XHIGH",
+    "Yunxi",
+    "YUNXI"
+  ]
+}

sdk/voicelive/azure-ai-voicelive/src/main/java/com/azure/ai/voicelive/models/AudioEchoCancellation.java

@@ -3,8 +3,8 @@
 // Code generated by Microsoft (R) TypeSpec Code Generator.
 package com.azure.ai.voicelive.models;
 
+import com.azure.core.annotation.Fluent;
 import com.azure.core.annotation.Generated;
-import com.azure.core.annotation.Immutable;
 import com.azure.json.JsonReader;
 import com.azure.json.JsonSerializable;
 import com.azure.json.JsonToken;
@@ -14,7 +14,7 @@
 /**
  * Echo cancellation configuration for server-side audio processing.
  */
-@Immutable
+@Fluent
 public final class AudioEchoCancellation implements JsonSerializable<AudioEchoCancellation> {
 
     /*
@@ -48,6 +48,9 @@ public String getType() {
     public JsonWriter toJson(JsonWriter jsonWriter) throws IOException {
         jsonWriter.writeStartObject();
         jsonWriter.writeStringField("type", this.type);
+        jsonWriter.writeStringField("reference_source",
+            this.referenceSource == null ? null : this.referenceSource.toString());
+        jsonWriter.writeNumberField("channels", this.channels);
         return jsonWriter.writeEndObject();
     }
 
@@ -67,9 +70,93 @@ public static AudioEchoCancellation fromJson(JsonReader jsonReader) throws IOExc
             while (reader.nextToken() != JsonToken.END_OBJECT) {
                 String fieldName = reader.getFieldName();
                 reader.nextToken();
-                reader.skipChildren();
+                if ("reference_source".equals(fieldName)) {
+                    deserializedAudioEchoCancellation.referenceSource
+                        = EchoCancellationReferenceSource.fromString(reader.getString());
+                } else if ("channels".equals(fieldName)) {
+                    deserializedAudioEchoCancellation.channels = reader.getNullable(JsonReader::getInt);
+                } else {
+                    reader.skipChildren();
+                }
             }
             return deserializedAudioEchoCancellation;
         });
     }
+
+    /*
+     * The source of the echo cancellation reference signal.
+     * - `server`: EC uses the internal TTS loopback as the reference signal (default, existing behavior).
+     * - `client`: EC uses the client-supplied reference channel (ch1 of stereo input). Internal TTS loopback is
+     * skipped.
+     */
+    @Generated
+    private EchoCancellationReferenceSource referenceSource;
+
+    /*
+     * Number of input audio channels.
+     * - `1`: Mono input (default).
+     * - `2`: Interleaved stereo input where channel 0 is the microphone signal and channel 1 is the echo reference
+     * signal.
+     * When set to 2, `reference_source` must be `client` and `input_audio_format` must be `pcm16`.
+     */
+    @Generated
+    private Integer channels;
+
+    /**
+     * Get the referenceSource property: The source of the echo cancellation reference signal.
+     * - `server`: EC uses the internal TTS loopback as the reference signal (default, existing behavior).
+     * - `client`: EC uses the client-supplied reference channel (ch1 of stereo input). Internal TTS loopback is
+     * skipped.
+     *
+     * @return the referenceSource value.
+     */
+    @Generated
+    public EchoCancellationReferenceSource getReferenceSource() {
+        return this.referenceSource;
+    }
+
+    /**
+     * Set the referenceSource property: The source of the echo cancellation reference signal.
+     * - `server`: EC uses the internal TTS loopback as the reference signal (default, existing behavior).
+     * - `client`: EC uses the client-supplied reference channel (ch1 of stereo input). Internal TTS loopback is
+     * skipped.
+     *
+     * @param referenceSource the referenceSource value to set.
+     * @return the AudioEchoCancellation object itself.
+     */
+    @Generated
+    public AudioEchoCancellation setReferenceSource(EchoCancellationReferenceSource referenceSource) {
+        this.referenceSource = referenceSource;
+        return this;
+    }
+
+    /**
+     * Get the channels property: Number of input audio channels.
+     * - `1`: Mono input (default).
+     * - `2`: Interleaved stereo input where channel 0 is the microphone signal and channel 1 is the echo reference
+     * signal.
+     * When set to 2, `reference_source` must be `client` and `input_audio_format` must be `pcm16`.
+     *
+     * @return the channels value.
+     */
+    @Generated
+    public Integer getChannels() {
+        return this.channels;
+    }
+
+    /**
+     * Set the channels property: Number of input audio channels.
+     * - `1`: Mono input (default).
+     * - `2`: Interleaved stereo input where channel 0 is the microphone signal and channel 1 is the echo reference
+     * signal.
+     * When set to 2, `reference_source` must be `client` and `input_audio_format` must be `pcm16`.
+     *
+     * @param channels the channels value to set.
+     * @return the AudioEchoCancellation object itself.
+     */
+    @Generated
+    public AudioEchoCancellation setChannels(Integer channels) {
+        this.channels = channels;
+        return this;
+    }
 }

sdk/voicelive/azure-ai-voicelive/src/main/java/com/azure/ai/voicelive/models/AzureRealtimeNativeVoice.java

@@ -0,0 +1,102 @@
+// Copyright (c) Microsoft Corporation. All rights reserved.
+// Licensed under the MIT License.
+// Code generated by Microsoft (R) TypeSpec Code Generator.
+package com.azure.ai.voicelive.models;
+
+import com.azure.core.annotation.Generated;
+import com.azure.core.annotation.Immutable;
+import com.azure.json.JsonReader;
+import com.azure.json.JsonSerializable;
+import com.azure.json.JsonToken;
+import com.azure.json.JsonWriter;
+import java.io.IOException;
+
+/**
+ * Azure realtime native voice configuration. These voices are natively
+ * supported by the `azure-realtime` model and offer higher quality speech
+ * synthesis than standard Azure voices. Only valid when using the
+ * `azure-realtime` model.
+ */
+@Immutable
+public final class AzureRealtimeNativeVoice implements JsonSerializable<AzureRealtimeNativeVoice> {
+
+    /*
+     * The type of the voice.
+     */
+    @Generated
+    private final String type = "azure-realtime-native";
+
+    /*
+     * The name of the Azure realtime native voice.
+     */
+    @Generated
+    private final AzureRealtimeNativeVoiceName name;
+
+    /**
+     * Creates an instance of AzureRealtimeNativeVoice class.
+     *
+     * @param name the name value to set.
+     */
+    @Generated
+    public AzureRealtimeNativeVoice(AzureRealtimeNativeVoiceName name) {
+        this.name = name;
+    }
+
+    /**
+     * Get the type property: The type of the voice.
+     *
+     * @return the type value.
+     */
+    @Generated
+    public String getType() {
+        return this.type;
+    }
+
+    /**
+     * Get the name property: The name of the Azure realtime native voice.
+     *
+     * @return the name value.
+     */
+    @Generated
+    public AzureRealtimeNativeVoiceName getName() {
+        return this.name;
+    }
+
+    /**
+     * {@inheritDoc}
+     */
+    @Generated
+    @Override
+    public JsonWriter toJson(JsonWriter jsonWriter) throws IOException {
+        jsonWriter.writeStartObject();
+        jsonWriter.writeStringField("type", this.type);
+        jsonWriter.writeStringField("name", this.name == null ? null : this.name.toString());
+        return jsonWriter.writeEndObject();
+    }
+
+    /**
+     * Reads an instance of AzureRealtimeNativeVoice from the JsonReader.
+     *
+     * @param jsonReader The JsonReader being read.
+     * @return An instance of AzureRealtimeNativeVoice if the JsonReader was pointing to an instance of it, or null if
+     * it was pointing to JSON null.
+     * @throws IllegalStateException If the deserialized JSON object was missing any required properties.
+     * @throws IOException If an error occurs while reading the AzureRealtimeNativeVoice.
+     */
+    @Generated
+    public static AzureRealtimeNativeVoice fromJson(JsonReader jsonReader) throws IOException {
+        return jsonReader.readObject(reader -> {
+            AzureRealtimeNativeVoiceName name = null;
+            while (reader.nextToken() != JsonToken.END_OBJECT) {
+                String fieldName = reader.getFieldName();
+                reader.nextToken();
+                if ("name".equals(fieldName)) {
+                    name = AzureRealtimeNativeVoiceName.fromString(reader.getString());
+                } else {
+                    reader.skipChildren();
+                }
+            }
+            return new AzureRealtimeNativeVoice(name);
+        });
+    }
+}