(inworld tts): add new model (#5646)

Author: tinalenguyen

livekit-agents/livekit/agents/inference/tts.py

@@ -52,6 +52,7 @@
 ]
 InworldModels = Literal[
     "inworld",
+    "inworld/inworld-tts-2",
     "inworld/inworld-tts-1.5-max",
     "inworld/inworld-tts-1.5-mini",
     "inworld/inworld-tts-1-max",

livekit-plugins/livekit-plugins-inworld/livekit/plugins/inworld/__init__.py

@@ -19,7 +19,15 @@
 """
 
 from .stt import STT, SpeechStream
-from .tts import TTS, ChunkedStream, Encoding, SynthesizeStream, TextNormalization, TimestampType
+from .tts import (
+    TTS,
+    ChunkedStream,
+    Encoding,
+    SynthesizeStream,
+    TextNormalization,
+    TimestampType,
+    TTSModels,
+)
 from .version import __version__
 
 __all__ = [
@@ -29,6 +37,7 @@
     "ChunkedStream",
     "SynthesizeStream",
     "Encoding",
+    "TTSModels",
     "TextNormalization",
     "TimestampType",
     "__version__",

livekit-plugins/livekit-plugins-inworld/livekit/plugins/inworld/tts.py

@@ -63,6 +63,7 @@
 DEFAULT_MAX_BUFFER_DELAY_MS = 3000
 NUM_CHANNELS = 1
 
+TTSModels = Literal["inworld-tts-2", "inworld-tts-1.5-max"]
 Encoding = Literal["LINEAR16", "PCM", "MP3", "OGG_OPUS", "FLAC"]
 TimestampType = Literal["TIMESTAMP_TYPE_UNSPECIFIED", "WORD", "CHARACTER"]
 _TextNormalizationStr = Literal["APPLY_TEXT_NORMALIZATION_UNSPECIFIED", "ON", "OFF"]
@@ -87,7 +88,7 @@ def _resolve_text_normalization(value: Any) -> _TextNormalizationStr:
 
 @dataclass
 class _TTSOptions:
-    model: str
+    model: TTSModels | str
     encoding: Encoding
     voice: str
     sample_rate: int
@@ -818,7 +819,7 @@ def __init__(
         *,
         api_key: NotGivenOr[str] = NOT_GIVEN,
         voice: NotGivenOr[str] = NOT_GIVEN,
-        model: NotGivenOr[str] = NOT_GIVEN,
+        model: NotGivenOr[TTSModels | str] = NOT_GIVEN,
         encoding: NotGivenOr[Encoding] = NOT_GIVEN,
         bit_rate: NotGivenOr[int] = NOT_GIVEN,
         sample_rate: NotGivenOr[int] = NOT_GIVEN,
@@ -976,7 +977,7 @@ def update_options(
         self,
         *,
         voice: NotGivenOr[str] = NOT_GIVEN,
-        model: NotGivenOr[str] = NOT_GIVEN,
+        model: NotGivenOr[TTSModels | str] = NOT_GIVEN,
         encoding: NotGivenOr[Encoding] = NOT_GIVEN,
         bit_rate: NotGivenOr[int] = NOT_GIVEN,
         sample_rate: NotGivenOr[int] = NOT_GIVEN,