add new room options

examples/voice_agents/basic_agent.py

@@ -11,12 +11,11 @@
     JobProcess,
     MetricsCollectedEvent,
     ModelSettings,
-    RoomInputOptions,
-    RoomOutputOptions,
     RunContext,
     WorkerOptions,
     cli,
     metrics,
+    room_io,
 )
 from livekit.agents.llm import function_tool
 from livekit.plugins import deepgram, openai, silero
@@ -123,11 +122,12 @@ async def log_usage():
     await session.start(
         agent=MyAgent(),
         room=ctx.room,
-        room_input_options=RoomInputOptions(
-            # uncomment to enable Krisp BVC noise cancellation
-            # noise_cancellation=noise_cancellation.BVC(),
+        room_options=room_io.RoomOptions(
+            audio_input=room_io.AudioInputOptions(
+                # uncomment to enable the Krisp BVC noise cancellation
+                # noise_cancellation=noise_cancellation.BVC(),
+            ),
         ),
-        room_output_options=RoomOutputOptions(transcription_enabled=True),
     )
 
 

livekit-agents/livekit/agents/__init__.py

@@ -75,6 +75,7 @@
     UserStateChangedEvent,
     avatar,
     io,
+    room_io,
 )
 from .voice.background_audio import AudioConfig, BackgroundAudioPlayer, BuiltinAudioClip, PlayHandle
 from .voice.room_io import RoomInputOptions, RoomIO, RoomOutputOptions
@@ -127,6 +128,7 @@ def __getattr__(name: str) -> typing.Any:
     "function_tool",
     "ChatContext",
     "ChatItem",
+    "room_io",
     "RoomIO",
     "RoomInputOptions",
     "RoomOutputOptions",

livekit-agents/livekit/agents/voice/agent_session.py

@@ -414,6 +414,8 @@ async def start(
         *,
         capture_run: Literal[True],
         room: NotGivenOr[rtc.Room] = NOT_GIVEN,
+        room_options: NotGivenOr[room_io.RoomOptions] = NOT_GIVEN,
+        # deprecated
         room_input_options: NotGivenOr[room_io.RoomInputOptions] = NOT_GIVEN,
         room_output_options: NotGivenOr[room_io.RoomOutputOptions] = NOT_GIVEN,
     ) -> RunResult: ...
@@ -425,6 +427,8 @@ async def start(
         *,
         capture_run: Literal[False] = False,
         room: NotGivenOr[rtc.Room] = NOT_GIVEN,
+        room_options: NotGivenOr[room_io.RoomOptions] = NOT_GIVEN,
+        # deprecated
         room_input_options: NotGivenOr[room_io.RoomInputOptions] = NOT_GIVEN,
         room_output_options: NotGivenOr[room_io.RoomOutputOptions] = NOT_GIVEN,
     ) -> None: ...
@@ -436,6 +440,8 @@ async def start(
         *,
         capture_run: bool = False,
         room: NotGivenOr[rtc.Room] = NOT_GIVEN,
+        room_options: NotGivenOr[room_io.RoomOptions] = NOT_GIVEN,
+        # deprecated
         room_input_options: NotGivenOr[room_io.RoomInputOptions] = NOT_GIVEN,
         room_output_options: NotGivenOr[room_io.RoomOutputOptions] = NOT_GIVEN,
     ) -> RunResult | None:
@@ -485,40 +491,35 @@ async def start(
                 tasks.append(asyncio.create_task(chat_cli.start(), name="_chat_cli_start"))
 
             elif is_given(room) and not self._room_io:
-                room_input_options = copy.copy(
-                    room_input_options or room_io.DEFAULT_ROOM_INPUT_OPTIONS
-                )
-                room_output_options = copy.copy(
-                    room_output_options or room_io.DEFAULT_ROOM_OUTPUT_OPTIONS
+                room_options = room_io.RoomOptions._ensure_options(
+                    room_options,
+                    room_input_options=room_input_options,
+                    room_output_options=room_output_options,
                 )
+                room_options = copy.copy(room_options)  # shadow copy is enough
 
                 if self.input.audio is not None:
-                    if room_input_options.audio_enabled:
+                    if room_options.audio_input:
                         logger.warning(
                             "RoomIO audio input is enabled but input.audio is already set, ignoring.."  # noqa: E501
                         )
-                    room_input_options.audio_enabled = False
+                    room_options.audio_input = False
 
                 if self.output.audio is not None:
-                    if room_output_options.audio_enabled:
+                    if room_options.audio_output:
                         logger.warning(
                             "RoomIO audio output is enabled but output.audio is already set, ignoring.."  # noqa: E501
                         )
-                    room_output_options.audio_enabled = False
+                    room_options.audio_output = False
 
                 if self.output.transcription is not None:
-                    if room_output_options.transcription_enabled:
+                    if room_options.text_output:
                         logger.warning(
                             "RoomIO transcription output is enabled but output.transcription is already set, ignoring.."  # noqa: E501
                         )
-                    room_output_options.transcription_enabled = False
+                    room_options.text_output = False
 
-                self._room_io = room_io.RoomIO(
-                    room=room,
-                    agent_session=self,
-                    input_options=room_input_options,
-                    output_options=room_output_options,
-                )
+                self._room_io = room_io.RoomIO(room=room, agent_session=self, options=room_options)
                 tasks.append(asyncio.create_task(self._room_io.start(), name="_room_io_start"))
 
             # session can be restarted, register the callbacks only once

livekit-agents/livekit/agents/voice/room_io/__init__.py

@@ -4,23 +4,31 @@
     _ParticipantStreamTranscriptionOutput,
     _ParticipantTranscriptionOutput,
 )
-from .room_io import (
-    DEFAULT_ROOM_INPUT_OPTIONS,
-    DEFAULT_ROOM_OUTPUT_OPTIONS,
+from .room_io import RoomIO
+from .types import (
+    AudioInputOptions,
+    AudioOutputOptions,
     RoomInputOptions,
-    RoomIO,
+    RoomOptions,
     RoomOutputOptions,
     TextInputEvent,
+    TextInputOptions,
+    TextOutputOptions,
+    VideoInputOptions,
 )
 
 __all__ = [
     "RoomIO",
-    "DEFAULT_ROOM_INPUT_OPTIONS",
-    "DEFAULT_ROOM_OUTPUT_OPTIONS",
+    "RoomOptions",
     "RoomInputOptions",
     "RoomOutputOptions",
     "ATTRIBUTE_PUBLISH_ON_BEHALF",
     "TextInputEvent",
+    "TextInputOptions",
+    "AudioInputOptions",
+    "AudioOutputOptions",
+    "TextOutputOptions",
+    "VideoInputOptions",
     "_ParticipantTranscriptionOutput",
     "_ParticipantAudioOutput",
     "_ParticipantStreamTranscriptionOutput",