livekit-plugins-google 1.2.1__tar.gz → 1.2.3__tar.gz

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/.gitignore

@@ -169,4 +169,7 @@ node_modules
 
 credentials.json
 pyrightconfig.json
-docs/
+docs/
+
+# Database files
+*.db

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: livekit-plugins-google
-Version: 1.2.1
+Version: 1.2.3
 Summary: Agent Framework plugin for services from Google Cloud
 Project-URL: Documentation, https://docs.livekit.io
 Project-URL: Website, https://livekit.io/
@@ -22,7 +22,7 @@ Requires-Dist: google-auth<3,>=2
 Requires-Dist: google-cloud-speech<3,>=2
 Requires-Dist: google-cloud-texttospeech<3,>=2.27
 Requires-Dist: google-genai>=v1.23.0
-Requires-Dist: livekit-agents>=1.2.1
+Requires-Dist: livekit-agents>=1.2.3
 Description-Content-Type: text/markdown
 
 # Google AI plugin for LiveKit Agents

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/livekit/plugins/google/beta/realtime/api_proto.py

@@ -9,6 +9,7 @@ LiveAPIModels = Literal[
     "gemini-2.0-flash-exp",
     # models supported on Gemini API
     "gemini-2.0-flash-live-001",
+    "gemini-live-2.5-flash-preview",
     "gemini-2.5-flash-preview-native-audio-dialog",
     "gemini-2.5-flash-exp-native-audio-thinking-dialog",
 ]

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/livekit/plugins/google/beta/realtime/realtime_api.py

@@ -8,6 +8,7 @@ import time
 import weakref
 from collections.abc import Iterator
 from dataclasses import dataclass, field
+from typing import Literal
 
 from google import genai
 from google.genai import types
@@ -542,7 +543,12 @@ class RealtimeSession(llm.RealtimeSession):
         self.start_user_activity()
 
     def truncate(
-        self, *, message_id: str, audio_end_ms: int, audio_transcript: NotGivenOr[str] = NOT_GIVEN
+        self,
+        *,
+        message_id: str,
+        modalities: list[Literal["text", "audio"]],
+        audio_end_ms: int,
+        audio_transcript: NotGivenOr[str] = NOT_GIVEN,
     ) -> None:
         logger.warning("truncate is not supported by the Google Realtime API.")
         pass
@@ -799,11 +805,16 @@ class RealtimeSession(llm.RealtimeSession):
         if not self._realtime_model.capabilities.audio_output:
             self._current_generation.audio_ch.close()
 
+        msg_modalities = asyncio.Future[list[Literal["text", "audio"]]]()
+        msg_modalities.set_result(
+            ["audio", "text"] if self._realtime_model.capabilities.audio_output else ["text"]
+        )
         self._current_generation.message_ch.send_nowait(
             llm.MessageGeneration(
                 message_id=response_id,
                 text_stream=self._current_generation.text_ch,
                 audio_stream=self._current_generation.audio_ch,
+                modalities=msg_modalities,
             )
         )
 
@@ -817,6 +828,10 @@ class RealtimeSession(llm.RealtimeSession):
             generation_event.user_initiated = True
             self._pending_generation_fut.set_result(generation_event)
             self._pending_generation_fut = None
+        else:
+            # emit input_speech_started event before starting an agent initiated generation
+            # to interrupt the previous audio playout if any
+            self._handle_input_speech_started()
 
         self.emit("generation_created", generation_event)
 
@@ -882,6 +897,9 @@ class RealtimeSession(llm.RealtimeSession):
         if not self._current_generation or self._current_generation._done:
             return
 
+        # emit input_speech_stopped event after the generation is done
+        self._handle_input_speech_stopped()
+
         gen = self._current_generation
 
         # The only way we'd know that the transcription is complete is by when they are
@@ -926,6 +944,12 @@ class RealtimeSession(llm.RealtimeSession):
     def _handle_input_speech_started(self) -> None:
         self.emit("input_speech_started", llm.InputSpeechStartedEvent())
 
+    def _handle_input_speech_stopped(self) -> None:
+        self.emit(
+            "input_speech_stopped",
+            llm.InputSpeechStoppedEvent(user_transcription_enabled=False),
+        )
+
     def _handle_tool_calls(self, tool_call: types.LiveServerToolCall) -> None:
         if not self._current_generation:
             logger.warning("received tool call but no active generation.")

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/livekit/plugins/google/version.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__version__ = "1.2.1"
+__version__ = "1.2.3"

{livekit_plugins_google-1.2.1 → livekit_plugins_google-1.2.3}/pyproject.toml

@@ -27,7 +27,7 @@ dependencies = [
     "google-cloud-speech >= 2, < 3",
     "google-cloud-texttospeech >= 2.27, < 3",
     "google-genai >= v1.23.0",
-    "livekit-agents>=1.2.1",
+    "livekit-agents>=1.2.3",
 ]
 
 [project.urls]