sarvamai 0.1.8rc6__py3-none-any.whl → 0.1.9a2__py3-none-any.whl

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,
@@ -189,6 +189,13 @@ __all__ = [
     "CompletionUsageParams",
     "ConfigMessage",
     "ConfigMessageParams",
+    "ConfigureConnection",
+    "ConfigureConnectionData",
+    "ConfigureConnectionDataOutputAudioBitrate",
+    "ConfigureConnectionDataParams",
+    "ConfigureConnectionDataSpeaker",
+    "ConfigureConnectionDataTargetLanguageCode",
+    "ConfigureConnectionParams",
     "CreateChatCompletionResponse",
     "CreateChatCompletionResponseParams",
     "DiarizedEntry",
@@ -213,13 +220,6 @@ __all__ = [
     "FlushSignalParams",
     "ForbiddenError",
     "Format",
-    "InitializeConnection",
-    "InitializeConnectionData",
-    "InitializeConnectionDataOutputAudioBitrate",
-    "InitializeConnectionDataParams",
-    "InitializeConnectionDataSpeaker",
-    "InitializeConnectionDataTargetLanguageCode",
-    "InitializeConnectionParams",
     "InternalServerError",
     "LanguageIdentificationResponse",
     "LanguageIdentificationResponseParams",

sarvamai/client.py

@@ -35,6 +35,9 @@ class SarvamAI:
 
 
     api_subscription_key : typing.Optional[str]
+    headers : typing.Optional[typing.Dict[str, str]]
+        Additional headers to send with every request.
+
     timeout : typing.Optional[float]
         The timeout to be used, in seconds, for requests. By default the timeout is 60 seconds, unless a custom httpx client is used, in which case this default is not enforced.
 
@@ -58,6 +61,7 @@ class SarvamAI:
         *,
         environment: SarvamAIEnvironment = SarvamAIEnvironment.PRODUCTION,
         api_subscription_key: typing.Optional[str] = os.getenv("SARVAM_API_KEY"),
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         timeout: typing.Optional[float] = None,
         follow_redirects: typing.Optional[bool] = True,
         httpx_client: typing.Optional[httpx.Client] = None,
@@ -72,6 +76,7 @@ class SarvamAI:
         self._client_wrapper = SyncClientWrapper(
             environment=environment,
             api_subscription_key=api_subscription_key,
+            headers=headers,
             httpx_client=httpx_client
             if httpx_client is not None
             else httpx.Client(timeout=_defaulted_timeout, follow_redirects=follow_redirects)
@@ -106,6 +111,9 @@ class AsyncSarvamAI:
 
 
     api_subscription_key : typing.Optional[str]
+    headers : typing.Optional[typing.Dict[str, str]]
+        Additional headers to send with every request.
+
     timeout : typing.Optional[float]
         The timeout to be used, in seconds, for requests. By default the timeout is 60 seconds, unless a custom httpx client is used, in which case this default is not enforced.
 
@@ -129,6 +137,7 @@ class AsyncSarvamAI:
         *,
         environment: SarvamAIEnvironment = SarvamAIEnvironment.PRODUCTION,
         api_subscription_key: typing.Optional[str] = os.getenv("SARVAM_API_KEY"),
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         timeout: typing.Optional[float] = None,
         follow_redirects: typing.Optional[bool] = True,
         httpx_client: typing.Optional[httpx.AsyncClient] = None,
@@ -143,6 +152,7 @@ class AsyncSarvamAI:
         self._client_wrapper = AsyncClientWrapper(
             environment=environment,
             api_subscription_key=api_subscription_key,
+            headers=headers,
             httpx_client=httpx_client
             if httpx_client is not None
             else httpx.AsyncClient(timeout=_defaulted_timeout, follow_redirects=follow_redirects)

sarvamai/core/client_wrapper.py

@@ -9,22 +9,32 @@ from .http_client import AsyncHttpClient, HttpClient
 
 class BaseClientWrapper:
     def __init__(
-        self, *, api_subscription_key: str, environment: SarvamAIEnvironment, timeout: typing.Optional[float] = None
+        self,
+        *,
+        api_subscription_key: str,
+        headers: typing.Optional[typing.Dict[str, str]] = None,
+        environment: SarvamAIEnvironment,
+        timeout: typing.Optional[float] = None,
     ):
         self.api_subscription_key = api_subscription_key
+        self._headers = headers
         self._environment = environment
         self._timeout = timeout
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.8rc6",
+            "User-Agent": "sarvamai/0.1.9a2",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.8rc6",
+            "X-Fern-SDK-Version": "0.1.9a2",
+            **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key
         return headers
 
+    def get_custom_headers(self) -> typing.Optional[typing.Dict[str, str]]:
+        return self._headers
+
     def get_environment(self) -> SarvamAIEnvironment:
         return self._environment
 
@@ -37,11 +47,14 @@ class SyncClientWrapper(BaseClientWrapper):
         self,
         *,
         api_subscription_key: str,
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         environment: SarvamAIEnvironment,
         timeout: typing.Optional[float] = None,
         httpx_client: httpx.Client,
     ):
-        super().__init__(api_subscription_key=api_subscription_key, environment=environment, timeout=timeout)
+        super().__init__(
+            api_subscription_key=api_subscription_key, headers=headers, environment=environment, timeout=timeout
+        )
         self.httpx_client = HttpClient(
             httpx_client=httpx_client, base_headers=self.get_headers, base_timeout=self.get_timeout
         )
@@ -52,11 +65,14 @@ class AsyncClientWrapper(BaseClientWrapper):
         self,
         *,
         api_subscription_key: str,
+        headers: typing.Optional[typing.Dict[str, str]] = None,
         environment: SarvamAIEnvironment,
         timeout: typing.Optional[float] = None,
         httpx_client: httpx.AsyncClient,
     ):
-        super().__init__(api_subscription_key=api_subscription_key, environment=environment, timeout=timeout)
+        super().__init__(
+            api_subscription_key=api_subscription_key, headers=headers, environment=environment, timeout=timeout
+        )
         self.httpx_client = AsyncHttpClient(
             httpx_client=httpx_client, base_headers=self.get_headers, base_timeout=self.get_timeout
         )

sarvamai/core/pydantic_utilities.py

@@ -59,9 +59,9 @@ class UniversalBaseModel(pydantic.BaseModel):
             protected_namespaces=(),
         )
 
-        @pydantic.model_serializer(mode="wrap", when_used="json")  # type: ignore[attr-defined]
-        def serialize_model(self, handler: pydantic.SerializerFunctionWrapHandler) -> Any:  # type: ignore[name-defined]
-            serialized = handler(self)
+        @pydantic.model_serializer(mode="plain", when_used="json")  # type: ignore[attr-defined]
+        def serialize_model(self) -> Any:  # type: ignore[name-defined]
+            serialized = self.model_dump()
             data = {k: serialize_datetime(v) if isinstance(v, dt.datetime) else v for k, v in serialized.items()}
             return data
 

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/requests/{initialize_connection.py → configure_connection.py}

@@ -3,10 +3,10 @@
 import typing
 
 import typing_extensions
-from .initialize_connection_data import InitializeConnectionDataParams
+from .configure_connection_data import ConfigureConnectionDataParams
 
 
-class InitializeConnectionParams(typing_extensions.TypedDict):
+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
@@ -15,4 +15,4 @@ class InitializeConnectionParams(typing_extensions.TypedDict):
     """
 
     type: typing.Literal["config"]
-    data: InitializeConnectionDataParams
+    data: ConfigureConnectionDataParams

sarvamai/requests/{initialize_connection_data.py → configure_connection_data.py}

@@ -3,18 +3,18 @@
 import typing
 
 import typing_extensions
-from ..types.initialize_connection_data_output_audio_bitrate import InitializeConnectionDataOutputAudioBitrate
-from ..types.initialize_connection_data_speaker import InitializeConnectionDataSpeaker
-from ..types.initialize_connection_data_target_language_code import InitializeConnectionDataTargetLanguageCode
+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 InitializeConnectionDataParams(typing_extensions.TypedDict):
-    target_language_code: InitializeConnectionDataTargetLanguageCode
+class ConfigureConnectionDataParams(typing_extensions.TypedDict):
+    target_language_code: ConfigureConnectionDataTargetLanguageCode
     """
     The language of the text is BCP-47 format
     """
 
-    speaker: InitializeConnectionDataSpeaker
+    speaker: ConfigureConnectionDataSpeaker
     """
     The speaker voice to be used for the output audio.
     
@@ -67,7 +67,7 @@ class InitializeConnectionDataParams(typing_extensions.TypedDict):
     Audio codec (currently supports MP3 only, optimized for real-time playback)
     """
 
-    output_audio_bitrate: typing_extensions.NotRequired[InitializeConnectionDataOutputAudioBitrate]
+    output_audio_bitrate: typing_extensions.NotRequired[ConfigureConnectionDataOutputAudioBitrate]
     """
     Audio bitrate (choose from 5 supported bitrate options)
     """

sarvamai/speech_to_text/client.py

@@ -61,14 +61,12 @@ class SpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2`
+            Note:- Default model is `saarika:v2.5`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio. This parameter is required to ensure accurate transcription.
-             For the `saarika:v1` model, this parameter is mandatory.
-             For the `saarika:v2` model, it is optional.
+            Specifies the language of the input audio.
+             For the `saarika:v2.5` model, it is optional.
             `unknown`: Use this when the language is not known; the API will detect it automatically.
-            Note:- that the `saarika:v1` model does not support `unknown` language code.
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -192,14 +190,12 @@ class AsyncSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2`
+            Note:- Default model is `saarika:v2.5`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio. This parameter is required to ensure accurate transcription.
-             For the `saarika:v1` model, this parameter is mandatory.
-             For the `saarika:v2` model, it is optional.
+            Specifies the language of the input audio.
+             For the `saarika:v2.5` model, it is optional.
             `unknown`: Use this when the language is not known; the API will detect it automatically.
-            Note:- that the `saarika:v1` model does not support `unknown` language code.
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.

sarvamai/speech_to_text/raw_client.py

@@ -60,14 +60,12 @@ class RawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2`
+            Note:- Default model is `saarika:v2.5`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio. This parameter is required to ensure accurate transcription.
-             For the `saarika:v1` model, this parameter is mandatory.
-             For the `saarika:v2` model, it is optional.
+            Specifies the language of the input audio.
+             For the `saarika:v2.5` model, it is optional.
             `unknown`: Use this when the language is not known; the API will detect it automatically.
-            Note:- that the `saarika:v1` model does not support `unknown` language code.
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -346,14 +344,12 @@ class AsyncRawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2`
+            Note:- Default model is `saarika:v2.5`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio. This parameter is required to ensure accurate transcription.
-             For the `saarika:v1` model, this parameter is mandatory.
-             For the `saarika:v2` model, it is optional.
+            Specifies the language of the input audio.
+             For the `saarika:v2.5` model, it is optional.
             `unknown`: Use this when the language is not known; the API will detect it automatically.
-            Note:- that the `saarika:v1` model does not support `unknown` language code.
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.

sarvamai/text/client.py

@@ -202,7 +202,7 @@ class TextClient:
         Parameters
         ----------
         input : str
-            The text input for language and script identification.
+            The text input for language and script identification. Max Input Limit is 1000 characters
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -534,7 +534,7 @@ class AsyncTextClient:
         Parameters
         ----------
         input : str
-            The text input for language and script identification.
+            The text input for language and script identification. Max Input Limit is 1000 characters
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.

sarvamai/text/raw_client.py

@@ -263,7 +263,7 @@ class RawTextClient:
         Parameters
         ----------
         input : str
-            The text input for language and script identification.
+            The text input for language and script identification. Max Input Limit is 1000 characters
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
@@ -773,7 +773,7 @@ class AsyncRawTextClient:
         Parameters
         ----------
         input : str
-            The text input for language and script identification.
+            The text input for language and script identification. Max Input Limit is 1000 characters
 
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.

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/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",

sarvamai/types/{initialize_connection.py → configure_connection.py}

@@ -4,10 +4,10 @@ import typing
 
 import pydantic
 from ..core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
-from .initialize_connection_data import InitializeConnectionData
+from .configure_connection_data import ConfigureConnectionData
 
 
-class InitializeConnection(UniversalBaseModel):
+class ConfigureConnection(UniversalBaseModel):
     """
     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
@@ -16,7 +16,7 @@ class InitializeConnection(UniversalBaseModel):
     """
 
     type: typing.Literal["config"] = "config"
-    data: InitializeConnectionData
+    data: ConfigureConnectionData
 
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2

sarvamai/types/{initialize_connection_data.py → configure_connection_data.py}

@@ -4,18 +4,18 @@ import typing
 
 import pydantic
 from ..core.pydantic_utilities import IS_PYDANTIC_V2, UniversalBaseModel
-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 .configure_connection_data_output_audio_bitrate import ConfigureConnectionDataOutputAudioBitrate
+from .configure_connection_data_speaker import ConfigureConnectionDataSpeaker
+from .configure_connection_data_target_language_code import ConfigureConnectionDataTargetLanguageCode
 
 
-class InitializeConnectionData(UniversalBaseModel):
-    target_language_code: InitializeConnectionDataTargetLanguageCode = pydantic.Field()
+class ConfigureConnectionData(UniversalBaseModel):
+    target_language_code: ConfigureConnectionDataTargetLanguageCode = pydantic.Field()
     """
     The language of the text is BCP-47 format
     """
 
-    speaker: InitializeConnectionDataSpeaker = pydantic.Field()
+    speaker: ConfigureConnectionDataSpeaker = pydantic.Field()
     """
     The speaker voice to be used for the output audio.
     
@@ -68,7 +68,7 @@ class InitializeConnectionData(UniversalBaseModel):
     Audio codec (currently supports MP3 only, optimized for real-time playback)
     """
 
-    output_audio_bitrate: typing.Optional[InitializeConnectionDataOutputAudioBitrate] = pydantic.Field(default=None)
+    output_audio_bitrate: typing.Optional[ConfigureConnectionDataOutputAudioBitrate] = pydantic.Field(default=None)
     """
     Audio bitrate (choose from 5 supported bitrate options)
     """

sarvamai/types/{initialize_connection_data_output_audio_bitrate.py → configure_connection_data_output_audio_bitrate.py}

@@ -2,6 +2,6 @@
 
 import typing
 
-InitializeConnectionDataOutputAudioBitrate = typing.Union[
+ConfigureConnectionDataOutputAudioBitrate = typing.Union[
     typing.Literal["32k", "64k", "96k", "128k", "192k"], typing.Any
 ]

sarvamai/types/{initialize_connection_data_speaker.py → configure_connection_data_speaker.py}

@@ -2,6 +2,6 @@
 
 import typing
 
-InitializeConnectionDataSpeaker = typing.Union[
+ConfigureConnectionDataSpeaker = typing.Union[
     typing.Literal["anushka", "abhilash", "manisha", "vidya", "arya", "karun", "hitesh"], typing.Any
 ]

sarvamai/types/{initialize_connection_data_target_language_code.py → configure_connection_data_target_language_code.py}

@@ -2,7 +2,7 @@
 
 import typing
 
-InitializeConnectionDataTargetLanguageCode = typing.Union[
+ConfigureConnectionDataTargetLanguageCode = typing.Union[
     typing.Literal["bn-IN", "en-IN", "gu-IN", "hi-IN", "kn-IN", "ml-IN", "mr-IN", "od-IN", "pa-IN", "ta-IN", "te-IN"],
     typing.Any,
 ]

sarvamai/types/speech_to_text_model.py

@@ -2,6 +2,4 @@
 
 import typing
 
-SpeechToTextModel = typing.Union[
-    typing.Literal["saarika:v1", "saarika:v2", "saarika:v2.5", "saarika:flash"], typing.Any
-]
+SpeechToTextModel = typing.Union[typing.Literal["saarika:v2", "saarika:v2.5"], typing.Any]

sarvamai/types/speech_to_text_translate_model.py

@@ -2,6 +2,4 @@
 
 import typing
 
-SpeechToTextTranslateModel = typing.Union[
-    typing.Literal["saaras:v1", "saaras:v2", "saaras:v2.5", "saaras:flash"], typing.Any
-]
+SpeechToTextTranslateModel = typing.Union[typing.Literal["saaras:v2", "saaras:v2.5"], typing.Any]

{sarvamai-0.1.8rc6.dist-info → sarvamai-0.1.9a2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.8rc6
+Version: 0.1.9a2
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers
@@ -20,9 +20,9 @@ Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Typing :: Typed
 Requires-Dist: httpx (>=0.21.2)
 Requires-Dist: pydantic (>=1.9.2)
-Requires-Dist: pydantic-core (>=2.18.2,<3.0.0)
+Requires-Dist: pydantic-core (>=2.18.2)
 Requires-Dist: typing_extensions (>=4.0.0)
-Requires-Dist: websockets (==12.0)
+Requires-Dist: websockets (>=12.0)
 Description-Content-Type: text/markdown
 
 # Sarvam Python Library

{sarvamai-0.1.8rc6.dist-info → sarvamai-0.1.9a2.dist-info}/RECORD

@@ -1,11 +1,11 @@
-sarvamai/__init__.py,sha256=PHOfArF05nQpPgHVlAeWbzuU3fj_a_1nb4KTPV5OuoE,8838
+sarvamai/__init__.py,sha256=xkQNmsHue7UaEN7PgNfH2ExBGBdyszDgg5omMkSnZMM,8824
 sarvamai/chat/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/chat/client.py,sha256=xOSj83Gr6Q7eY2qUeATiuXYQqBqWqSCQlIEopK5fKus,11022
 sarvamai/chat/raw_client.py,sha256=A2kRuZcVWlJhyYCD7YKgqNkZEp3cYa1731KhRkhirU0,17885
-sarvamai/client.py,sha256=aI1sw5LVGMjgukgZLDlUmA17ecK1yGsQxH-W_JiCrco,7177
+sarvamai/client.py,sha256=xaDMuAdwCIKkq0iyY4Gp6M4alMcGOErkFOlZMI1bvNA,7581
 sarvamai/core/__init__.py,sha256=YE2CtXeASe1RAbaI39twKWYKCuT4tW5is9HWHhJjR_g,1653
 sarvamai/core/api_error.py,sha256=44vPoTyWN59gonCIZMdzw7M1uspygiLnr3GNFOoVL2Q,614
-sarvamai/core/client_wrapper.py,sha256=XLVb0UOmyM8ir1Ht42b7D13ojMz0QetLyGKaImhbMcI,2080
+sarvamai/core/client_wrapper.py,sha256=Hb9s60KyyAif9RPeTsPXfvf92h4mj27Okb0GsvLcFys,2568
 sarvamai/core/datetime_utils.py,sha256=nBys2IsYrhPdszxGKCNRPSOCwa-5DWOHG95FB8G9PKo,1047
 sarvamai/core/events.py,sha256=j7VWXgMpOsjCXdzY22wIhI7Q-v5InZ4WchRzA88x_Sk,856
 sarvamai/core/file.py,sha256=d4NNbX8XvXP32z8KpK2Xovv33nFfruIrpz0QWxlgpZk,2663
@@ -13,7 +13,7 @@ sarvamai/core/force_multipart.py,sha256=awxh5MtcRYe74ehY8U76jzv6fYM_w_D3Rur7KQQz
 sarvamai/core/http_client.py,sha256=QurkBvCZZz2Z1d8znp4M2YbOXebBUPcPXRhPIS84Wvk,21214
 sarvamai/core/http_response.py,sha256=4uOAtXXFTyFXHLXeQWSfQST9PGcOCRAdHVgGTxdyg84,1334
 sarvamai/core/jsonable_encoder.py,sha256=hGgcEEeX11sqxxsll7h15pO3pTNVxk_n79Kcn0laoWA,3655
-sarvamai/core/pydantic_utilities.py,sha256=HxbbISfaP1XBvzbIkc0ZcF_GHKd9BfYsJAcFh9B126k,10787
+sarvamai/core/pydantic_utilities.py,sha256=cqBsZZCJC3PJacYdbJgyVLEhviH0BT6isZu5BGLzsII,10743
 sarvamai/core/query_encoder.py,sha256=ekulqNd0j8TgD7ox-Qbz7liqX8-KP9blvT9DsRCenYM,2144
 sarvamai/core/remove_none_from_dict.py,sha256=EU9SGgYidWq7SexuJbNs4-PZ-5Bl3Vppd864mS6vQZw,342
 sarvamai/core/request_options.py,sha256=h0QUNCFVdCW_7GclVySCAY2w4NhtXVBUCmHgmzaxpcg,1681
@@ -28,7 +28,7 @@ sarvamai/errors/too_many_requests_error.py,sha256=Dl-_pfpboXJh-OtSbRaPQOB-UXvpVO
 sarvamai/errors/unprocessable_entity_error.py,sha256=JqxtzIhvjkpQDqbT9Q-go1n-gyv9PsYqq0ng_ZYyBMo,347
 sarvamai/play.py,sha256=4fh86zy8g8IPU2O8yPBY7QxXQOivv_nWQvPQsOa1arw,2183
 sarvamai/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-sarvamai/requests/__init__.py,sha256=UmEj75u8yYiZYL0M0QVJoCuS8oduaH82eD3bmP-6myg,4484
+sarvamai/requests/__init__.py,sha256=AYpixS_3RLRqMW4U2OfeACJjD3u3NnethhdYr_V-P9M,4478
 sarvamai/requests/audio_data.py,sha256=QI3SK5aiAg2yJ-m3l9CxOkONnH3CCKMFCl9kAdMs19o,410
 sarvamai/requests/audio_message.py,sha256=ZBeogjGE6YFXXM-0g8zq9SoizDk21reR0YXSB-0fMjg,214
 sarvamai/requests/audio_output.py,sha256=BnoX345rwoWgaMaj24u_19-SjmPV0xt7vlFEEDKRw20,280
@@ -41,6 +41,8 @@ sarvamai/requests/chat_completion_response_message.py,sha256=JFazj4zK-nj_wjdvNLD
 sarvamai/requests/choice.py,sha256=uulX4MZUoThEMcD3a80o_3V5YpnpqN8DfPaNZWVz-1o,867
 sarvamai/requests/completion_usage.py,sha256=LbZV-RxcxKdCAYqhCiaRtSFF3VwMJq71A989Z1rm-I8,428
 sarvamai/requests/config_message.py,sha256=EpYioGvDhCXDMvGH7Q1F7448zJzoHmlkQ1owoNGbWAw,383
+sarvamai/requests/configure_connection.py,sha256=a-foQtLxArL4CulvKEdeebbRqmS1GRmko3MZdnHVPEk,716
+sarvamai/requests/configure_connection_data.py,sha256=Niil2OrVBzQEtmWFn1JC-StLVp6WzzRIsu2i_M_8_44,2908
 sarvamai/requests/create_chat_completion_response.py,sha256=TqS9u5_WVWMok_NreT4TeOsLJQeybPkbJm45Q0Zxw30,857
 sarvamai/requests/diarized_entry.py,sha256=gbXB4D_r5_Q8gs1arRKjxPeFcYg16dVDLcg2VhxmKQA,462
 sarvamai/requests/diarized_transcript.py,sha256=X-znuJ45oqwXzVyJumBHSqVGLz6JnoYFZmluQlEpEAw,323
@@ -51,8 +53,6 @@ sarvamai/requests/error_response.py,sha256=A8j12JQ7JJkUcnt26k2M9uwXXkwyT-LNqG3BO
 sarvamai/requests/error_response_data.py,sha256=l9tGTykaKZ8pKxdw9RKitpW49kKcs4aGibH7rKG2v7w,461
 sarvamai/requests/events_data.py,sha256=3seSash8DysPUWX6mKPzoEzWZlsrK4Tann2GFSbQjZg,286
 sarvamai/requests/flush_signal.py,sha256=Aj_PzphMNcHMMOaxvTi1uQ5y36ZTtKEsUGCprbWIOvw,406
-sarvamai/requests/initialize_connection.py,sha256=H4SMx6-TkJ_oyDtQkUpbS8H-njItwNDfHkl9bppLuwk,720
-sarvamai/requests/initialize_connection_data.py,sha256=NEvOWOO5UmIH3qxXsfBgjH2sLnG54HrR9x1AdYfBgCE,2918
 sarvamai/requests/language_identification_response.py,sha256=BdS5U9Gic-71vb--ph6HGvd2hGNKDXERC7yrn8vFcvI,1098
 sarvamai/requests/ping_signal.py,sha256=TSgmfz2k4X1L6TzvX8u2SKZ6XQY3bSf7nPZf8mUViaM,343
 sarvamai/requests/send_text.py,sha256=DWzbNgeNN2xSIYgk2zEisgLqjwq5oleqJVHrtOnIqbE,267
@@ -72,8 +72,8 @@ sarvamai/requests/transcription_metrics.py,sha256=FDclX2Z9Z3azrDXxtZW8xbkxxWMZQX
 sarvamai/requests/translation_response.py,sha256=8iwQeZB1purHY757bIQI-n9QeVRBItaAVcBJ_la-k1Y,414
 sarvamai/requests/transliteration_response.py,sha256=KqRkqnegLmt7LjdVxjRePX6RoqaLm64KFGZ6q7mXyfw,426
 sarvamai/speech_to_text/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
-sarvamai/speech_to_text/client.py,sha256=E76V1BZ236AN37cvm7c-NTRTnDleiYQ4_hKHE3wfevY,11435
-sarvamai/speech_to_text/raw_client.py,sha256=rencVGZ5cVX1eHhYoVMhcrlkqVmpI4aAGTrSC7kcYUI,25690
+sarvamai/speech_to_text/client.py,sha256=lp2G2fI9SUbeOBBE1S5tjcp-Xb8wIhAuVadLKwXveh8,11003
+sarvamai/speech_to_text/raw_client.py,sha256=-7CcxWpfJGqwsQ1efO53-NryG0pYzr1YKb6d9PVQb9k,25258
 sarvamai/speech_to_text_streaming/__init__.py,sha256=q7QygMmZCHJ-4FMhhL_6XNV_dsqlIFRCO1iSxoyxaaY,437
 sarvamai/speech_to_text_streaming/client.py,sha256=WdkzZxKMdnQ2hHv9hzJlfSNggRJLKFljRiC7695Jcog,8224
 sarvamai/speech_to_text_streaming/raw_client.py,sha256=7zcgVw7CXA2TySkJKQkS-NdnJOMRudmV_m4NaigICN0,7405
@@ -92,16 +92,16 @@ sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_strea
 sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_model.py,sha256=6B8VxkpJG_pNprCSctseDtJb_ULVdKrPaeENkQ6Jvjg,187
 sarvamai/speech_to_text_translate_streaming/types/speech_to_text_translate_streaming_vad_signals.py,sha256=EV3xd9qyKMnMvA9rO-qFDDIac4b84roBu7n-maaPxG8,181
 sarvamai/text/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
-sarvamai/text/client.py,sha256=CMwDVtMsPor08e8F9a7yhHtgrdLzd1__samwm4S9CGM,30525
-sarvamai/text/raw_client.py,sha256=lQ7bV9aVqxjwEUHMPEZ4x0_Xs036_yFArMK9rnYT4ZI,48625
+sarvamai/text/client.py,sha256=2kA0Gxfi-r52zMQdqRRD811014alzlHB_FANkp3Kn_c,30595
+sarvamai/text/raw_client.py,sha256=7xYmJA50kTKy_gj8tkAPckKp2djHB37zOdm0_icbMb8,48695
 sarvamai/text_to_speech/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/text_to_speech/client.py,sha256=aVvwdGTfgVUekALLpdyxKNsGUZQ2Ee2OZBLx6WP6g_E,8842
 sarvamai/text_to_speech/raw_client.py,sha256=3Zu6HN_FOY683Vm-EN-OL7YAbLsftjJlFm5OyRGNtYc,14780
 sarvamai/text_to_speech_streaming/__init__.py,sha256=_VhToAyIt_5axN6CLJwtxg3-CO7THa_23pbUzqhXJa4,85
 sarvamai/text_to_speech_streaming/client.py,sha256=geTF5xy-batzO12XVt0sPw_XJCi7-m2sDFK_B7SL7qc,6088
 sarvamai/text_to_speech_streaming/raw_client.py,sha256=asOcNw1WAViOiXDVWH4sxWSXGVoLwAOh9vUtq_xralA,5269
-sarvamai/text_to_speech_streaming/socket_client.py,sha256=qZpHFC1z1c6T_FaGOW6CPtWE15glmqClmdc5Iixa7FY,10429
-sarvamai/types/__init__.py,sha256=_5KKlTtdY4kOQ3go9T60EPmSA53AOFGbF5VJOkcnros,6320
+sarvamai/text_to_speech_streaming/socket_client.py,sha256=NEcijnvjuNcWfzqpBi-xWsXVkL0NPq6EGAkEjnaq9hw,13909
+sarvamai/types/__init__.py,sha256=cEfGVQMYlbz13iS9v83_CB9160Cky4JOImmxM30fGhg,6305
 sarvamai/types/audio_data.py,sha256=rgOukLkLNJ_HBBVE2g5dfEL2CWjRoGiMvCtpq0qTB1Y,829
 sarvamai/types/audio_message.py,sha256=sB4EgkWkWJzipYXobkmM9AYZTTZtCpg_ySKssUeznUE,560
 sarvamai/types/audio_output.py,sha256=Eq-YUZa1mSDwt7bax2c4Vv2gBlyM_JBJWzHhTAhFSko,621
@@ -114,6 +114,11 @@ sarvamai/types/chat_completion_response_message.py,sha256=wz935eBnCkSIl0I0qMxBuH
 sarvamai/types/choice.py,sha256=uXBCsjWP9VK3XWQWZUeI4EnU10w0G9nAfKn2tJZvxko,1244
 sarvamai/types/completion_usage.py,sha256=xYQGlQUbKqsksuV73H-1ajjfT5M7w47eLfdWXSlrI5M,843
 sarvamai/types/config_message.py,sha256=sGrT-qYTRqLVfIo5nRUuRlqPtPVmiAkUAnaMtlmQYCU,778
+sarvamai/types/configure_connection.py,sha256=SnSNk02gQqP8e4VB4y88jjeFQ4ClpImjGLn2ANI8cZ4,1058
+sarvamai/types/configure_connection_data.py,sha256=brMO-Z1TDq3oTJ22m1icBkkmnd9k67p_DzecnMcqNko,3421
+sarvamai/types/configure_connection_data_output_audio_bitrate.py,sha256=h00YvKLxsZC8L3__rH4XH53nN_GY40UElW1EjysCwUs,208
+sarvamai/types/configure_connection_data_speaker.py,sha256=SzyAiK5LynXwb9KniaO2qoOLY-II3-PMZbRuIsQ9shw,230
+sarvamai/types/configure_connection_data_target_language_code.py,sha256=jrU1EblAtDYbybUO1KUkHhevmlSBj2AQxX13ii3QhAQ,275
 sarvamai/types/create_chat_completion_response.py,sha256=4nEzeWzHGW1_BmRAtOuGsbRZ0ojNgnzJSMUFyYuYviw,1285
 sarvamai/types/diarized_entry.py,sha256=kf9DLrcoMHZdTKNCAaF0z46q_iAe7CE-DFP4CNrZGTw,896
 sarvamai/types/diarized_transcript.py,sha256=a491XmALLE7AQcByaaOYTew0BZoFTlewEMHLMJyj-Js,669
@@ -127,11 +132,6 @@ sarvamai/types/events_data.py,sha256=hDSOyODc8-lmpduJIQkps9kHlUZKYXGw3lETi8irHt0
 sarvamai/types/finish_reason.py,sha256=PBWtBNkX4FMaODmlUehpF6qLB5uH_zR-Mw3M4uhIB6U,209
 sarvamai/types/flush_signal.py,sha256=N7MJWb658KoxRpFN9cIbyQGY45zZcg8YCou3E1v--9o,759
 sarvamai/types/format.py,sha256=57LicD0XLqW4D1QEnZWsWGifzRy1GV9P5utKPXLoxtg,144
-sarvamai/types/initialize_connection.py,sha256=cXHmWJ1GrD1JX9fDMpQzUSRBXWTDAPekoEYtESRtQos,1062
-sarvamai/types/initialize_connection_data.py,sha256=vfHqhN29dg6IwjYMsDKgoAaoYwi6JAljPOQTrCx5CEM,3431
-sarvamai/types/initialize_connection_data_output_audio_bitrate.py,sha256=Wl2b6KR_shf6FzsBr1RJxBeVduS6AtO8skwQqOaffFc,209
-sarvamai/types/initialize_connection_data_speaker.py,sha256=pdLyERHk0NtAKj-7lvebElF7QR1ac2T2jZ-keJsWba8,231
-sarvamai/types/initialize_connection_data_target_language_code.py,sha256=iEuRFQJPsXO6lCwbNOoB7BzvoLCPIXL7UbIOKOMIjPs,276
 sarvamai/types/language_identification_response.py,sha256=jG4ZQ6KQHCiEDqC51OniOwiRdW14Fbz22bbTsUDp_kc,1483
 sarvamai/types/numerals_format.py,sha256=xg3lYiHcnzyFwuwRcaIteJLH_Pz6pJ9n9kTlYPEnCBU,165
 sarvamai/types/ping_signal.py,sha256=cE53FRIXlc8bSo18z6jlAnOh6DhZEMX36huWEX6X3-A,695
@@ -143,13 +143,13 @@ sarvamai/types/send_text.py,sha256=kgST6V5NuURzgBpuiDi8fVwDg768ttDoeY9k1dSSb1Y,6
 sarvamai/types/send_text_data.py,sha256=H8yfcvd4gvyN34RrZ9i4qQvieednNBhL7i7isX4asuY,519
 sarvamai/types/speech_sample_rate.py,sha256=Hfi79KL2Y1W7OIvvrfWnt7EUvmU5i7bxYvXivrY_aUA,88
 sarvamai/types/speech_to_text_language.py,sha256=cq8FBOX0DfYB3v8jgNteQtHeJcqWqzKWJVyYGwwo_w0,279
-sarvamai/types/speech_to_text_model.py,sha256=cYANeE-1Kjdpju97bLdQa8GuGqMBn0hVK4McFVqUGT0,208
+sarvamai/types/speech_to_text_model.py,sha256=oSbEfOyRdA2seM2XF-JZzA0EO0Vj8pha8jFWRzlmFlM,171
 sarvamai/types/speech_to_text_response.py,sha256=iWRGEJeHUFIOxeEhoCQu68njeA6lcqXbT2czV-O8Wx0,1438
 sarvamai/types/speech_to_text_response_data.py,sha256=gbxZTBSjbN3ZIa10K6tWPYtymcpnQTFIaUnXkOmsmD4,322
 sarvamai/types/speech_to_text_streaming_response.py,sha256=z6tVAHbVK9lC3w3lac__LEUfO8AAzEilkeGlaLskTtc,687
 sarvamai/types/speech_to_text_transcription_data.py,sha256=EqwPAPSi98PwARaTj-ufzFUSHyN-NPoPla5vi_KERrU,1297
 sarvamai/types/speech_to_text_translate_language.py,sha256=yikNM-roIumVG-eqBWss93wLGudZdLPwd0i3VcXH5zo,263
-sarvamai/types/speech_to_text_translate_model.py,sha256=hxKnhnelhH5Ah8QO5SymTJMnDw_WWAhGDJnutoNk3qw,213
+sarvamai/types/speech_to_text_translate_model.py,sha256=GkN6IHD94we3hWCWtTMFT_ot_5roLCauHMLIj9rLesQ,178
 sarvamai/types/speech_to_text_translate_response.py,sha256=Z5Na7IQW2ok3TP21xd-jKkwioplEKfonNIMhoJQKkVw,1278
 sarvamai/types/speech_to_text_translate_response_data.py,sha256=_NlLVp7oQU3em_4E47QVbIP9nromPE07Z9HtMpY1lrU,359
 sarvamai/types/speech_to_text_translate_streaming_response.py,sha256=J6h3AGdAJxpODFs30bR-e6OaWKa__oVhwv_TrbPSO98,724
@@ -173,6 +173,6 @@ sarvamai/types/transliterate_mode.py,sha256=1jSEMlGcoLkWuk12TgoOpSgwifa4rThGKZ1h
 sarvamai/types/transliterate_source_language.py,sha256=bSY9wJszF0sg-Cgg6F-YcWC8ly1mIlj9rqa15-jBtx8,283
 sarvamai/types/transliteration_response.py,sha256=yt-lzTbDeJ_ZL4I8kQa6oESxA9ebeJJY7LfFHpdEsmM,815
 sarvamai/version.py,sha256=Qkp3Ee9YH-O9RTix90e0i7iNrFAGN-QDt2AFwGA4n8k,75
-sarvamai-0.1.8rc6.dist-info/METADATA,sha256=tSU-7crW7SJuPf0lcP4nigyvM2VN1rkBTe7KzaxmnGw,26760
-sarvamai-0.1.8rc6.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-sarvamai-0.1.8rc6.dist-info/RECORD,,
+sarvamai-0.1.9a2.dist-info/METADATA,sha256=TwDKldURWAuQ_H1tggRwDDS-VUlW2o4z6e7_lsvfQuE,26752
+sarvamai-0.1.9a2.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+sarvamai-0.1.9a2.dist-info/RECORD,,