sarvamai 0.1.8rc5__tar.gz → 0.1.8rc7__tar.gz

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.8rc5
+Version: 0.1.8rc7
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/pyproject.toml

@@ -3,7 +3,7 @@ name = "sarvamai"
 
 [tool.poetry]
 name = "sarvamai"
-version = "0.1.8rc5"
+version = "0.1.8rc7"
 description = ""
 readme = "README.md"
 authors = []

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/__init__.py

@@ -18,6 +18,11 @@ from .types import (
     Choice,
     CompletionUsage,
     ConfigMessage,
+    ConfigureConnection,
+    ConfigureConnectionData,
+    ConfigureConnectionDataOutputAudioBitrate,
+    ConfigureConnectionDataSpeaker,
+    ConfigureConnectionDataTargetLanguageCode,
     CreateChatCompletionResponse,
     DiarizedEntry,
     DiarizedTranscript,
@@ -31,11 +36,6 @@ from .types import (
     FinishReason,
     FlushSignal,
     Format,
-    InitializeConnection,
-    InitializeConnectionData,
-    InitializeConnectionDataOutputAudioBitrate,
-    InitializeConnectionDataSpeaker,
-    InitializeConnectionDataTargetLanguageCode,
     LanguageIdentificationResponse,
     NumeralsFormat,
     PingSignal,
@@ -112,6 +112,8 @@ from .requests import (
     ChoiceParams,
     CompletionUsageParams,
     ConfigMessageParams,
+    ConfigureConnectionDataParams,
+    ConfigureConnectionParams,
     CreateChatCompletionResponseParams,
     DiarizedEntryParams,
     DiarizedTranscriptParams,
@@ -122,8 +124,6 @@ from .requests import (
     ErrorResponseParams,
     EventsDataParams,
     FlushSignalParams,
-    InitializeConnectionDataParams,
-    InitializeConnectionParams,
     LanguageIdentificationResponseParams,
     PingSignalParams,
     SendTextDataParams,
@@ -154,7 +154,6 @@ from .speech_to_text_translate_streaming import (
     SpeechToTextTranslateStreamingModel,
     SpeechToTextTranslateStreamingVadSignals,
 )
-from .text_to_speech_streaming import TextToSpeechStreamingModel
 from .version import __version__
 
 __all__ = [
@@ -190,6 +189,13 @@ __all__ = [
     "CompletionUsageParams",
     "ConfigMessage",
     "ConfigMessageParams",
+    "ConfigureConnection",
+    "ConfigureConnectionData",
+    "ConfigureConnectionDataOutputAudioBitrate",
+    "ConfigureConnectionDataParams",
+    "ConfigureConnectionDataSpeaker",
+    "ConfigureConnectionDataTargetLanguageCode",
+    "ConfigureConnectionParams",
     "CreateChatCompletionResponse",
     "CreateChatCompletionResponseParams",
     "DiarizedEntry",
@@ -214,13 +220,6 @@ __all__ = [
     "FlushSignalParams",
     "ForbiddenError",
     "Format",
-    "InitializeConnection",
-    "InitializeConnectionData",
-    "InitializeConnectionDataOutputAudioBitrate",
-    "InitializeConnectionDataParams",
-    "InitializeConnectionDataSpeaker",
-    "InitializeConnectionDataTargetLanguageCode",
-    "InitializeConnectionParams",
     "InternalServerError",
     "LanguageIdentificationResponse",
     "LanguageIdentificationResponseParams",
@@ -274,7 +273,6 @@ __all__ = [
     "TextToSpeechResponse",
     "TextToSpeechResponseParams",
     "TextToSpeechSpeaker",
-    "TextToSpeechStreamingModel",
     "TimestampsModel",
     "TimestampsModelParams",
     "TooManyRequestsError",

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/core/client_wrapper.py

@@ -17,10 +17,10 @@ class BaseClientWrapper:
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.8rc5",
+            "User-Agent": "sarvamai/0.1.8rc7",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.8rc5",
+            "X-Fern-SDK-Version": "0.1.8rc7",
         }
         headers["api-subscription-key"] = self.api_subscription_key
         return headers

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/requests/__init__.py

@@ -19,6 +19,8 @@ from .chat_completion_response_message import ChatCompletionResponseMessageParam
 from .choice import ChoiceParams
 from .completion_usage import CompletionUsageParams
 from .config_message import ConfigMessageParams
+from .configure_connection import ConfigureConnectionParams
+from .configure_connection_data import ConfigureConnectionDataParams
 from .create_chat_completion_response import CreateChatCompletionResponseParams
 from .diarized_entry import DiarizedEntryParams
 from .diarized_transcript import DiarizedTranscriptParams
@@ -29,8 +31,6 @@ from .error_response import ErrorResponseParams
 from .error_response_data import ErrorResponseDataParams
 from .events_data import EventsDataParams
 from .flush_signal import FlushSignalParams
-from .initialize_connection import InitializeConnectionParams
-from .initialize_connection_data import InitializeConnectionDataParams
 from .language_identification_response import LanguageIdentificationResponseParams
 from .ping_signal import PingSignalParams
 from .send_text import SendTextParams
@@ -66,6 +66,8 @@ __all__ = [
     "ChoiceParams",
     "CompletionUsageParams",
     "ConfigMessageParams",
+    "ConfigureConnectionDataParams",
+    "ConfigureConnectionParams",
     "CreateChatCompletionResponseParams",
     "DiarizedEntryParams",
     "DiarizedTranscriptParams",
@@ -76,8 +78,6 @@ __all__ = [
     "ErrorResponseParams",
     "EventsDataParams",
     "FlushSignalParams",
-    "InitializeConnectionDataParams",
-    "InitializeConnectionParams",
     "LanguageIdentificationResponseParams",
     "PingSignalParams",
     "SendTextDataParams",

sarvamai-0.1.8rc7/src/sarvamai/requests/configure_connection.py

@@ -0,0 +1,18 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import typing_extensions
+from .configure_connection_data import ConfigureConnectionDataParams
+
+
+class ConfigureConnectionParams(typing_extensions.TypedDict):
+    """
+    Configuration message required as the first message after establishing the WebSocket connection.
+    This initializes TTS parameters and can be updated at any time during the WebSocket lifecycle
+    by sending a new config message. When a config update is sent, any text currently in the buffer
+    will be automatically flushed and processed before applying the new configuration.
+    """
+
+    type: typing.Literal["config"]
+    data: ConfigureConnectionDataParams

sarvamai-0.1.8rc7/src/sarvamai/requests/configure_connection_data.py

@@ -0,0 +1,83 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import typing_extensions
+from ..types.configure_connection_data_output_audio_bitrate import ConfigureConnectionDataOutputAudioBitrate
+from ..types.configure_connection_data_speaker import ConfigureConnectionDataSpeaker
+from ..types.configure_connection_data_target_language_code import ConfigureConnectionDataTargetLanguageCode
+
+
+class ConfigureConnectionDataParams(typing_extensions.TypedDict):
+    target_language_code: ConfigureConnectionDataTargetLanguageCode
+    """
+    The language of the text is BCP-47 format
+    """
+
+    speaker: ConfigureConnectionDataSpeaker
+    """
+    The speaker voice to be used for the output audio.
+    
+    **Default:** Anushka
+    
+    **Model Compatibility (Speakers compatible with respective model):**
+    - **bulbul:v2:**
+      - Female: Anushka, Manisha, Vidya, Arya
+      - Male: Abhilash, Karun, Hitesh
+    
+    **Note:** Speaker selection must match the chosen model version.
+    """
+
+    pitch: typing_extensions.NotRequired[float]
+    """
+    Controls the pitch of the audio. Lower values result in a deeper voice, 
+    while higher values make it sharper. The suitable range is between -0.75 
+    and 0.75. Default is 0.0.
+    """
+
+    pace: typing_extensions.NotRequired[float]
+    """
+    Controls the speed of the audio. Lower values result in slower speech, 
+    while higher values make it faster. The suitable range is between 0.5 
+    and 2.0. Default is 1.0.
+    """
+
+    loudness: typing_extensions.NotRequired[float]
+    """
+    Controls the loudness of the audio. Lower values result in quieter audio, 
+    while higher values make it louder. The suitable range is between 0.3 
+    and 3.0. Default is 1.0.
+    """
+
+    speech_sample_rate: typing_extensions.NotRequired[int]
+    """
+    Specifies the sample rate of the output audio. Supported values are 
+    8000, 16000, 22050, 24000 Hz. If not provided, the default is 22050 Hz.
+    """
+
+    enable_preprocessing: typing_extensions.NotRequired[bool]
+    """
+    Controls whether normalization of English words and numeric entities 
+    (e.g., numbers, dates) is performed. Set to true for better handling 
+    of mixed-language text. Default is false.
+    """
+
+    output_audio_codec: typing_extensions.NotRequired[typing.Literal["mp3"]]
+    """
+    Audio codec (currently supports MP3 only, optimized for real-time playback)
+    """
+
+    output_audio_bitrate: typing_extensions.NotRequired[ConfigureConnectionDataOutputAudioBitrate]
+    """
+    Audio bitrate (choose from 5 supported bitrate options)
+    """
+
+    min_buffer_size: typing_extensions.NotRequired[int]
+    """
+    Minimum character length that triggers buffer flushing for TTS model processing
+    """
+
+    max_chunk_length: typing_extensions.NotRequired[int]
+    """
+    Maximum length for sentence splitting (adjust based on content length)
+    """

sarvamai-0.1.8rc7/src/sarvamai/requests/flush_signal.py

@@ -0,0 +1,14 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+import typing_extensions
+
+
+class FlushSignalParams(typing_extensions.TypedDict):
+    """
+    Forces the text buffer to process immediately, regardless of the min_buffer_size threshold.
+    Use this when you need to process remaining text that hasn't reached the minimum buffer size.
+    """
+
+    type: typing.Literal["flush"]

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/requests/ping_signal.py

@@ -6,4 +6,9 @@ import typing_extensions
 
 
 class PingSignalParams(typing_extensions.TypedDict):
+    """
+    Send ping signal to keep the WebSocket connection alive. The connection automatically
+    closes after one minute of inactivity.
+    """
+
     type: typing.Literal["ping"]

sarvamai-0.1.8rc7/src/sarvamai/text_to_speech_streaming/__init__.py

@@ -0,0 +1,4 @@
+# This file was auto-generated by Fern from our API Definition.
+
+# isort: skip_file
+

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/text_to_speech_streaming/client.py

@@ -11,7 +11,6 @@ from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .raw_client import AsyncRawTextToSpeechStreamingClient, RawTextToSpeechStreamingClient
 from .socket_client import AsyncTextToSpeechStreamingSocketClient, TextToSpeechStreamingSocketClient
-from .types.text_to_speech_streaming_model import TextToSpeechStreamingModel
 
 
 class TextToSpeechStreamingClient:
@@ -33,7 +32,7 @@ class TextToSpeechStreamingClient:
     def connect(
         self,
         *,
-        model: typing.Optional[TextToSpeechStreamingModel] = None,
+        model: typing.Optional[typing.Literal["bulbul:v2"]] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[TextToSpeechStreamingSocketClient]:
@@ -43,7 +42,7 @@ class TextToSpeechStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[TextToSpeechStreamingModel]
+        model : typing.Optional[typing.Literal["bulbul:v2"]]
             Text to speech model to use
 
         api_subscription_key : typing.Optional[str]
@@ -103,7 +102,7 @@ class AsyncTextToSpeechStreamingClient:
     async def connect(
         self,
         *,
-        model: typing.Optional[TextToSpeechStreamingModel] = None,
+        model: typing.Optional[typing.Literal["bulbul:v2"]] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncTextToSpeechStreamingSocketClient]:
@@ -113,7 +112,7 @@ class AsyncTextToSpeechStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[TextToSpeechStreamingModel]
+        model : typing.Optional[typing.Literal["bulbul:v2"]]
             Text to speech model to use
 
         api_subscription_key : typing.Optional[str]

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/text_to_speech_streaming/raw_client.py

@@ -10,7 +10,6 @@ from ..core.api_error import ApiError
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from .socket_client import AsyncTextToSpeechStreamingSocketClient, TextToSpeechStreamingSocketClient
-from .types.text_to_speech_streaming_model import TextToSpeechStreamingModel
 
 
 class RawTextToSpeechStreamingClient:
@@ -21,7 +20,7 @@ class RawTextToSpeechStreamingClient:
     def connect(
         self,
         *,
-        model: typing.Optional[TextToSpeechStreamingModel] = None,
+        model: typing.Optional[typing.Literal["bulbul:v2"]] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[TextToSpeechStreamingSocketClient]:
@@ -31,7 +30,7 @@ class RawTextToSpeechStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[TextToSpeechStreamingModel]
+        model : typing.Optional[typing.Literal["bulbul:v2"]]
             Text to speech model to use
 
         api_subscription_key : typing.Optional[str]
@@ -80,7 +79,7 @@ class AsyncRawTextToSpeechStreamingClient:
     async def connect(
         self,
         *,
-        model: typing.Optional[TextToSpeechStreamingModel] = None,
+        model: typing.Optional[typing.Literal["bulbul:v2"]] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncTextToSpeechStreamingSocketClient]:
@@ -90,7 +89,7 @@ class AsyncRawTextToSpeechStreamingClient:
 
         Parameters
         ----------
-        model : typing.Optional[TextToSpeechStreamingModel]
+        model : typing.Optional[typing.Literal["bulbul:v2"]]
             Text to speech model to use
 
         api_subscription_key : typing.Optional[str]

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/text_to_speech_streaming/socket_client.py

@@ -10,8 +10,8 @@ from ..core.pydantic_utilities import parse_obj_as
 from ..types.audio_output import AudioOutput
 from ..types.flush_signal import FlushSignal
 from ..types.error_response import ErrorResponse
-from ..types.initialize_connection import InitializeConnection
-from ..types.initialize_connection_data import InitializeConnectionData
+from ..types.configure_connection import ConfigureConnection
+from ..types.configure_connection_data import ConfigureConnectionData
 from ..types.ping_signal import PingSignal
 from ..types.send_text import SendText
 from ..types.send_text_data import SendTextData
@@ -54,10 +54,10 @@ class AsyncTextToSpeechStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def initialize_connection(
+    async def configure(
         self,
         target_language_code: str,
-        speaker: str,
+        speaker: str = "anushka",
         pitch: float = 0.0,
         pace: float = 1.0,
         loudness: float = 1.0,
@@ -69,21 +69,35 @@ class AsyncTextToSpeechStreamingSocketClient(EventEmitterMixin):
         max_chunk_length: int = 150,
     ) -> None:
         """
-        Initialize the TTS connection with configuration parameters.
+        Configuration message required as the first message after establishing the WebSocket connection.
+        This initializes TTS parameters and can be updated at any time during the WebSocket lifecycle
+        by sending a new config message. When a config update is sent, any text currently in the buffer
+        will be automatically flushed and processed before applying the new configuration.
 
-        :param target_language_code: Target language code (e.g., 'hi-IN')
-        :param speaker: Voice speaker name (e.g., 'meera', 'arvind')
-        :param pitch: Voice pitch adjustment (-1.0 to 1.0, default: 0.0)
-        :param pace: Speech pace (0.3 to 3.0, default: 1.0)
-        :param loudness: Voice loudness (0.1 to 3.0, default: 1.0)
-        :param speech_sample_rate: Audio sample rate, default: 22050
-        :param enable_preprocessing: Enable text preprocessing, default: False
-        :param output_audio_codec: Audio codec, default: 'mp3'
-        :param output_audio_bitrate: Audio bitrate, default: '128k'
-        :param min_buffer_size: Minimum buffer size, default: 50
-        :param max_chunk_length: Maximum chunk length, default: 150
-        """
-        data = InitializeConnectionData(
+        :param target_language_code: The language of the text is BCP-47 format
+        :param speaker: The speaker voice to be used for the output audio. Default: Anushka.
+            Model Compatibility (bulbul:v2): Female: Anushka, Manisha, Vidya, Arya;
+            Male: Abhilash, Karun, Hitesh
+        :param pitch: Controls the pitch of the audio. Lower values result in a deeper voice,
+            while higher values make it sharper. The suitable range is between -0.75
+            and 0.75. Default is 0.0.
+        :param pace: Controls the speed of the audio. Lower values result in slower speech,
+            while higher values make it faster. The suitable range is between 0.5
+            and 2.0. Default is 1.0.
+        :param loudness: Controls the loudness of the audio. Lower values result in quieter audio,
+            while higher values make it louder. The suitable range is between 0.3
+            and 3.0. Default is 1.0.
+        :param speech_sample_rate: Specifies the sample rate of the output audio. Supported values are
+            8000, 16000, 22050, 24000 Hz. If not provided, the default is 22050 Hz.
+        :param enable_preprocessing: Controls whether normalization of English words and numeric entities
+            (e.g., numbers, dates) is performed. Set to true for better handling
+            of mixed-language text. Default is false.
+        :param output_audio_codec: Audio codec (currently supports MP3 only, optimized for real-time playback)
+        :param output_audio_bitrate: Audio bitrate (choose from 5 supported bitrate options)
+        :param min_buffer_size: Minimum character length that triggers buffer flushing for TTS model processing
+        :param max_chunk_length: Maximum length for sentence splitting (adjust based on content length)
+        """
+        data = ConfigureConnectionData(
             target_language_code=target_language_code,
             speaker=speaker,
             pitch=pitch,
@@ -96,14 +110,16 @@ class AsyncTextToSpeechStreamingSocketClient(EventEmitterMixin):
             min_buffer_size=min_buffer_size,
             max_chunk_length=max_chunk_length,
         )
-        message = InitializeConnection(data=data)
+        message = ConfigureConnection(data=data)
         await self._send_model(message)
 
     async def convert(self, text: str) -> None:
         """
-        Send text to be converted to speech.
+        Send text to be converted to speech. Text length should be 1-2500 characters.
+        Recommended: <500 characters for optimal streaming performance.
+        Real-time endpoints perform better with longer character counts.
 
-        :param text: Text to be synthesized (1-2500 characters)
+        :param text: Text to be synthesized (1-2500 characters, recommended <500)
         """
         data = SendTextData(text=text)
         message = SendText(data=data)
@@ -111,15 +127,16 @@ class AsyncTextToSpeechStreamingSocketClient(EventEmitterMixin):
 
     async def flush(self) -> None:
         """
-        Signal to flush the buffer and finalize audio output.
-        This indicates the end of text input.
+        Forces the text buffer to process immediately, regardless of the min_buffer_size threshold.
+        Use this when you need to process remaining text that hasn't reached the minimum buffer size.
         """
         message = FlushSignal()
         await self._send_model(message)
 
     async def ping(self) -> None:
         """
-        Send ping signal to keep the WebSocket connection alive.
+        Send ping signal to keep the WebSocket connection alive. The connection automatically
+        closes after one minute of inactivity.
         """
         message = PingSignal()
         await self._send_model(message)
@@ -182,10 +199,10 @@ class TextToSpeechStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def initialize_connection(
+    def configure(
         self,
         target_language_code: str,
-        speaker: str,
+        speaker: str = "anushka",
         pitch: float = 0.0,
         pace: float = 1.0,
         loudness: float = 1.0,
@@ -197,21 +214,35 @@ class TextToSpeechStreamingSocketClient(EventEmitterMixin):
         max_chunk_length: int = 150,
     ) -> None:
         """
-        Initialize the TTS connection with configuration parameters.
+        Configuration message required as the first message after establishing the WebSocket connection.
+        This initializes TTS parameters and can be updated at any time during the WebSocket lifecycle
+        by sending a new config message. When a config update is sent, any text currently in the buffer
+        will be automatically flushed and processed before applying the new configuration.
 
-        :param target_language_code: Target language code (e.g., 'hi-IN')
-        :param speaker: Voice speaker name (e.g., 'meera', 'arvind')
-        :param pitch: Voice pitch adjustment (-1.0 to 1.0, default: 0.0)
-        :param pace: Speech pace (0.3 to 3.0, default: 1.0)
-        :param loudness: Voice loudness (0.1 to 3.0, default: 1.0)
-        :param speech_sample_rate: Audio sample rate, default: 22050
-        :param enable_preprocessing: Enable text preprocessing, default: False
-        :param output_audio_codec: Audio codec, default: 'mp3'
-        :param output_audio_bitrate: Audio bitrate, default: '128k'
-        :param min_buffer_size: Minimum buffer size, default: 50
-        :param max_chunk_length: Maximum chunk length, default: 150
-        """
-        data = InitializeConnectionData(
+        :param target_language_code: The language of the text is BCP-47 format
+        :param speaker: The speaker voice to be used for the output audio. Default: Anushka.
+            Model Compatibility (bulbul:v2): Female: Anushka, Manisha, Vidya, Arya;
+            Male: Abhilash, Karun, Hitesh
+        :param pitch: Controls the pitch of the audio. Lower values result in a deeper voice,
+            while higher values make it sharper. The suitable range is between -0.75
+            and 0.75. Default is 0.0.
+        :param pace: Controls the speed of the audio. Lower values result in slower speech,
+            while higher values make it faster. The suitable range is between 0.5
+            and 2.0. Default is 1.0.
+        :param loudness: Controls the loudness of the audio. Lower values result in quieter audio,
+            while higher values make it louder. The suitable range is between 0.3
+            and 3.0. Default is 1.0.
+        :param speech_sample_rate: Specifies the sample rate of the output audio. Supported values are
+            8000, 16000, 22050, 24000 Hz. If not provided, the default is 22050 Hz.
+        :param enable_preprocessing: Controls whether normalization of English words and numeric entities
+            (e.g., numbers, dates) is performed. Set to true for better handling
+            of mixed-language text. Default is false.
+        :param output_audio_codec: Audio codec (currently supports MP3 only, optimized for real-time playback)
+        :param output_audio_bitrate: Audio bitrate (choose from 5 supported bitrate options)
+        :param min_buffer_size: Minimum character length that triggers buffer flushing for TTS model processing
+        :param max_chunk_length: Maximum length for sentence splitting (adjust based on content length)
+        """
+        data = ConfigureConnectionData(
             target_language_code=target_language_code,
             speaker=speaker,
             pitch=pitch,
@@ -224,14 +255,16 @@ class TextToSpeechStreamingSocketClient(EventEmitterMixin):
             min_buffer_size=min_buffer_size,
             max_chunk_length=max_chunk_length,
         )
-        message = InitializeConnection(data=data)
+        message = ConfigureConnection(data=data)
         self._send_model(message)
 
     def convert(self, text: str) -> None:
         """
-        Send text to be converted to speech.
+        Send text to be converted to speech. Text length should be 1-2500 characters.
+        Recommended: <500 characters for optimal streaming performance.
+        Real-time endpoints perform better with longer character counts.
 
-        :param text: Text to be synthesized (1-2500 characters)
+        :param text: Text to be synthesized (1-2500 characters, recommended <500)
         """
         data = SendTextData(text=text)
         message = SendText(data=data)
@@ -239,15 +272,16 @@ class TextToSpeechStreamingSocketClient(EventEmitterMixin):
 
     def flush(self) -> None:
         """
-        Signal to flush the buffer and finalize audio output.
-        This indicates the end of text input.
+        Forces the text buffer to process immediately, regardless of the min_buffer_size threshold.
+        Use this when you need to process remaining text that hasn't reached the minimum buffer size.
         """
         message = FlushSignal()
         self._send_model(message)
 
     def ping(self) -> None:
         """
-        Send ping signal to keep the WebSocket connection alive.
+        Send ping signal to keep the WebSocket connection alive. The connection automatically
+        closes after one minute of inactivity.
         """
         message = PingSignal()
         self._send_model(message)

{sarvamai-0.1.8rc5 → sarvamai-0.1.8rc7}/src/sarvamai/types/__init__.py

@@ -19,6 +19,11 @@ from .chat_completion_response_message import ChatCompletionResponseMessage
 from .choice import Choice
 from .completion_usage import CompletionUsage
 from .config_message import ConfigMessage
+from .configure_connection import ConfigureConnection
+from .configure_connection_data import ConfigureConnectionData
+from .configure_connection_data_output_audio_bitrate import ConfigureConnectionDataOutputAudioBitrate
+from .configure_connection_data_speaker import ConfigureConnectionDataSpeaker
+from .configure_connection_data_target_language_code import ConfigureConnectionDataTargetLanguageCode
 from .create_chat_completion_response import CreateChatCompletionResponse
 from .diarized_entry import DiarizedEntry
 from .diarized_transcript import DiarizedTranscript
@@ -32,11 +37,6 @@ from .events_data import EventsData
 from .finish_reason import FinishReason
 from .flush_signal import FlushSignal
 from .format import Format
-from .initialize_connection import InitializeConnection
-from .initialize_connection_data import InitializeConnectionData
-from .initialize_connection_data_output_audio_bitrate import InitializeConnectionDataOutputAudioBitrate
-from .initialize_connection_data_speaker import InitializeConnectionDataSpeaker
-from .initialize_connection_data_target_language_code import InitializeConnectionDataTargetLanguageCode
 from .language_identification_response import LanguageIdentificationResponse
 from .numerals_format import NumeralsFormat
 from .ping_signal import PingSignal
@@ -94,6 +94,11 @@ __all__ = [
     "Choice",
     "CompletionUsage",
     "ConfigMessage",
+    "ConfigureConnection",
+    "ConfigureConnectionData",
+    "ConfigureConnectionDataOutputAudioBitrate",
+    "ConfigureConnectionDataSpeaker",
+    "ConfigureConnectionDataTargetLanguageCode",
     "CreateChatCompletionResponse",
     "DiarizedEntry",
     "DiarizedTranscript",
@@ -107,11 +112,6 @@ __all__ = [
     "FinishReason",
     "FlushSignal",
     "Format",
-    "InitializeConnection",
-    "InitializeConnectionData",
-    "InitializeConnectionDataOutputAudioBitrate",
-    "InitializeConnectionDataSpeaker",
-    "InitializeConnectionDataTargetLanguageCode",
     "LanguageIdentificationResponse",
     "NumeralsFormat",
     "PingSignal",