cartesia 2.0.0b8__py3-none-any.whl → 2.0.3__py3-none-any.whl

cartesia/__init__.py

@@ -1,7 +1,8 @@
 # This file was auto-generated by Fern from our API Definition.
 
-from . import api_status, datasets, embedding, infill, tts, voice_changer, voices
+from . import api_status, auth, datasets, embedding, infill, tts, voice_changer, voices
 from .api_status import ApiInfo, ApiInfoParams
+from .auth import TokenGrant, TokenGrantParams, TokenRequest, TokenRequestParams, TokenResponse, TokenResponseParams
 from .client import AsyncCartesia, Cartesia
 from .datasets import (
     CreateDatasetRequest,
@@ -28,6 +29,7 @@ from .tts import (
     FlushId,
     GenerationRequest,
     GenerationRequestParams,
+    ModelSpeed,
     Mp3OutputFormat,
     Mp3OutputFormatParams,
     NaturalSpecifier,
@@ -200,6 +202,7 @@ __all__ = [
     "MixVoiceSpecifierParams",
     "MixVoicesRequest",
     "MixVoicesRequestParams",
+    "ModelSpeed",
     "Mp3OutputFormat",
     "Mp3OutputFormatParams",
     "NaturalSpecifier",
@@ -233,6 +236,12 @@ __all__ = [
     "StreamingResponse_Error",
     "StreamingResponse_ErrorParams",
     "SupportedLanguage",
+    "TokenGrant",
+    "TokenGrantParams",
+    "TokenRequest",
+    "TokenRequestParams",
+    "TokenResponse",
+    "TokenResponseParams",
     "TtsRequest",
     "TtsRequestEmbeddingSpecifier",
     "TtsRequestEmbeddingSpecifierParams",
@@ -294,6 +303,7 @@ __all__ = [
     "WordTimestampsParams",
     "__version__",
     "api_status",
+    "auth",
     "datasets",
     "embedding",
     "infill",

cartesia/auth/__init__.py

@@ -0,0 +1,13 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from .types import TokenGrant, TokenRequest, TokenResponse
+from .requests import TokenGrantParams, TokenRequestParams, TokenResponseParams
+
+__all__ = [
+    "TokenGrant",
+    "TokenGrantParams",
+    "TokenRequest",
+    "TokenRequestParams",
+    "TokenResponse",
+    "TokenResponseParams",
+]

cartesia/auth/client.py

@@ -0,0 +1,159 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+from ..core.client_wrapper import SyncClientWrapper
+from .requests.token_grant import TokenGrantParams
+from ..core.request_options import RequestOptions
+from .types.token_response import TokenResponse
+from ..core.serialization import convert_and_respect_annotation_metadata
+from ..core.pydantic_utilities import parse_obj_as
+from json.decoder import JSONDecodeError
+from ..core.api_error import ApiError
+from ..core.client_wrapper import AsyncClientWrapper
+
+# this is used as the default value for optional parameters
+OMIT = typing.cast(typing.Any, ...)
+
+
+class AuthClient:
+    def __init__(self, *, client_wrapper: SyncClientWrapper):
+        self._client_wrapper = client_wrapper
+
+    def access_token(
+        self,
+        *,
+        grants: TokenGrantParams,
+        expires_in: typing.Optional[int] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> TokenResponse:
+        """
+        Generates a new Access Token for the client. These tokens are short-lived and should be used to make requests to the API from authenticated clients.
+
+        Parameters
+        ----------
+        grants : TokenGrantParams
+            The permissions to be granted via the token.
+
+        expires_in : typing.Optional[int]
+            The number of seconds the token will be valid for since the time of generation. The maximum is 1 hour (3600 seconds).
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        TokenResponse
+
+        Examples
+        --------
+        from cartesia import Cartesia
+
+        client = Cartesia(
+            api_key="YOUR_API_KEY",
+        )
+        client.auth.access_token(
+            grants={"tts": True},
+            expires_in=60,
+        )
+        """
+        _response = self._client_wrapper.httpx_client.request(
+            "access-token",
+            method="POST",
+            json={
+                "grants": convert_and_respect_annotation_metadata(
+                    object_=grants, annotation=TokenGrantParams, direction="write"
+                ),
+                "expires_in": expires_in,
+            },
+            request_options=request_options,
+            omit=OMIT,
+        )
+        try:
+            if 200 <= _response.status_code < 300:
+                return typing.cast(
+                    TokenResponse,
+                    parse_obj_as(
+                        type_=TokenResponse,  # type: ignore
+                        object_=_response.json(),
+                    ),
+                )
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)
+
+
+class AsyncAuthClient:
+    def __init__(self, *, client_wrapper: AsyncClientWrapper):
+        self._client_wrapper = client_wrapper
+
+    async def access_token(
+        self,
+        *,
+        grants: TokenGrantParams,
+        expires_in: typing.Optional[int] = OMIT,
+        request_options: typing.Optional[RequestOptions] = None,
+    ) -> TokenResponse:
+        """
+        Generates a new Access Token for the client. These tokens are short-lived and should be used to make requests to the API from authenticated clients.
+
+        Parameters
+        ----------
+        grants : TokenGrantParams
+            The permissions to be granted via the token.
+
+        expires_in : typing.Optional[int]
+            The number of seconds the token will be valid for since the time of generation. The maximum is 1 hour (3600 seconds).
+
+        request_options : typing.Optional[RequestOptions]
+            Request-specific configuration.
+
+        Returns
+        -------
+        TokenResponse
+
+        Examples
+        --------
+        import asyncio
+
+        from cartesia import AsyncCartesia
+
+        client = AsyncCartesia(
+            api_key="YOUR_API_KEY",
+        )
+
+
+        async def main() -> None:
+            await client.auth.access_token(
+                grants={"tts": True},
+                expires_in=60,
+            )
+
+
+        asyncio.run(main())
+        """
+        _response = await self._client_wrapper.httpx_client.request(
+            "access-token",
+            method="POST",
+            json={
+                "grants": convert_and_respect_annotation_metadata(
+                    object_=grants, annotation=TokenGrantParams, direction="write"
+                ),
+                "expires_in": expires_in,
+            },
+            request_options=request_options,
+            omit=OMIT,
+        )
+        try:
+            if 200 <= _response.status_code < 300:
+                return typing.cast(
+                    TokenResponse,
+                    parse_obj_as(
+                        type_=TokenResponse,  # type: ignore
+                        object_=_response.json(),
+                    ),
+                )
+            _response_json = _response.json()
+        except JSONDecodeError:
+            raise ApiError(status_code=_response.status_code, body=_response.text)
+        raise ApiError(status_code=_response.status_code, body=_response_json)

cartesia/auth/requests/__init__.py

@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from .token_grant import TokenGrantParams
+from .token_request import TokenRequestParams
+from .token_response import TokenResponseParams
+
+__all__ = ["TokenGrantParams", "TokenRequestParams", "TokenResponseParams"]

cartesia/auth/requests/token_grant.py

@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing_extensions
+
+
+class TokenGrantParams(typing_extensions.TypedDict):
+    tts: bool
+    """
+    The `tts` grant allows the token to be used to access any TTS endpoint.
+    """

cartesia/auth/requests/token_request.py

@@ -0,0 +1,17 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing_extensions
+from .token_grant import TokenGrantParams
+import typing_extensions
+
+
+class TokenRequestParams(typing_extensions.TypedDict):
+    grants: TokenGrantParams
+    """
+    The permissions to be granted via the token.
+    """
+
+    expires_in: typing_extensions.NotRequired[int]
+    """
+    The number of seconds the token will be valid for since the time of generation. The maximum is 1 hour (3600 seconds).
+    """

cartesia/auth/requests/token_response.py

@@ -0,0 +1,10 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing_extensions
+
+
+class TokenResponseParams(typing_extensions.TypedDict):
+    token: str
+    """
+    The generated Access Token.
+    """

cartesia/auth/types/__init__.py

@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from .token_grant import TokenGrant
+from .token_request import TokenRequest
+from .token_response import TokenResponse
+
+__all__ = ["TokenGrant", "TokenRequest", "TokenResponse"]

cartesia/auth/types/token_grant.py

@@ -0,0 +1,22 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from ...core.pydantic_utilities import UniversalBaseModel
+import pydantic
+from ...core.pydantic_utilities import IS_PYDANTIC_V2
+import typing
+
+
+class TokenGrant(UniversalBaseModel):
+    tts: bool = pydantic.Field()
+    """
+    The `tts` grant allows the token to be used to access any TTS endpoint.
+    """
+
+    if IS_PYDANTIC_V2:
+        model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
+    else:
+
+        class Config:
+            frozen = True
+            smart_union = True
+            extra = pydantic.Extra.allow

cartesia/auth/types/token_request.py

@@ -0,0 +1,28 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from ...core.pydantic_utilities import UniversalBaseModel
+from .token_grant import TokenGrant
+import pydantic
+import typing
+from ...core.pydantic_utilities import IS_PYDANTIC_V2
+
+
+class TokenRequest(UniversalBaseModel):
+    grants: TokenGrant = pydantic.Field()
+    """
+    The permissions to be granted via the token.
+    """
+
+    expires_in: typing.Optional[int] = pydantic.Field(default=None)
+    """
+    The number of seconds the token will be valid for since the time of generation. The maximum is 1 hour (3600 seconds).
+    """
+
+    if IS_PYDANTIC_V2:
+        model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
+    else:
+
+        class Config:
+            frozen = True
+            smart_union = True
+            extra = pydantic.Extra.allow

cartesia/auth/types/token_response.py

@@ -0,0 +1,22 @@
+# This file was auto-generated by Fern from our API Definition.
+
+from ...core.pydantic_utilities import UniversalBaseModel
+import pydantic
+from ...core.pydantic_utilities import IS_PYDANTIC_V2
+import typing
+
+
+class TokenResponse(UniversalBaseModel):
+    token: str = pydantic.Field()
+    """
+    The generated Access Token.
+    """
+
+    if IS_PYDANTIC_V2:
+        model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
+    else:
+
+        class Config:
+            frozen = True
+            smart_union = True
+            extra = pydantic.Extra.allow

cartesia/base_client.py

@@ -5,12 +5,14 @@ from .environment import CartesiaEnvironment
 import httpx
 from .core.client_wrapper import SyncClientWrapper
 from .api_status.client import ApiStatusClient
+from .auth.client import AuthClient
 from .infill.client import InfillClient
 from .tts.client import TtsClient
 from .voice_changer.client import VoiceChangerClient
 from .voices.client import VoicesClient
 from .core.client_wrapper import AsyncClientWrapper
 from .api_status.client import AsyncApiStatusClient
+from .auth.client import AsyncAuthClient
 from .infill.client import AsyncInfillClient
 from .tts.client import AsyncTtsClient
 from .voice_changer.client import AsyncVoiceChangerClient
@@ -76,6 +78,7 @@ class BaseCartesia:
             timeout=_defaulted_timeout,
         )
         self.api_status = ApiStatusClient(client_wrapper=self._client_wrapper)
+        self.auth = AuthClient(client_wrapper=self._client_wrapper)
         self.infill = InfillClient(client_wrapper=self._client_wrapper)
         self.tts = TtsClient(client_wrapper=self._client_wrapper)
         self.voice_changer = VoiceChangerClient(client_wrapper=self._client_wrapper)
@@ -141,6 +144,7 @@ class AsyncBaseCartesia:
             timeout=_defaulted_timeout,
         )
         self.api_status = AsyncApiStatusClient(client_wrapper=self._client_wrapper)
+        self.auth = AsyncAuthClient(client_wrapper=self._client_wrapper)
         self.infill = AsyncInfillClient(client_wrapper=self._client_wrapper)
         self.tts = AsyncTtsClient(client_wrapper=self._client_wrapper)
         self.voice_changer = AsyncVoiceChangerClient(client_wrapper=self._client_wrapper)

cartesia/core/client_wrapper.py

@@ -16,7 +16,7 @@ class BaseClientWrapper:
         headers: typing.Dict[str, str] = {
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "cartesia",
-            "X-Fern-SDK-Version": "2.0.0b8",
+            "X-Fern-SDK-Version": "2.0.3",
         }
         headers["X-API-Key"] = self.api_key
         headers["Cartesia-Version"] = "2024-11-13"

cartesia/tts/__init__.py

@@ -7,6 +7,7 @@ from .types import (
     Emotion,
     FlushId,
     GenerationRequest,
+    ModelSpeed,
     Mp3OutputFormat,
     NaturalSpecifier,
     NumericalSpecifier,
@@ -94,6 +95,7 @@ __all__ = [
     "FlushId",
     "GenerationRequest",
     "GenerationRequestParams",
+    "ModelSpeed",
     "Mp3OutputFormat",
     "Mp3OutputFormatParams",
     "NaturalSpecifier",

cartesia/tts/_async_websocket.py

@@ -71,6 +71,7 @@ class _AsyncTTSContext:
         add_phoneme_timestamps: bool = False,
         use_original_timestamps: bool = False,
         continue_: bool = False,
+        max_buffer_delay_ms: Optional[int] = None,
         flush: bool = False,
     ) -> None:
         """Send audio generation requests to the WebSocket. The response can be received using the `receive` method.
@@ -111,6 +112,8 @@ class _AsyncTTSContext:
             request_body["use_original_timestamps"] = use_original_timestamps
         if continue_:
             request_body["continue"] = continue_
+        if max_buffer_delay_ms:
+            request_body["max_buffer_delay_ms"] = max_buffer_delay_ms
         if flush:
             request_body["flush"] = flush
 

cartesia/tts/_websocket.py

@@ -63,6 +63,7 @@ class _TTSContext:
         output_format: OutputFormatParams,
         voice: TtsRequestVoiceSpecifierParams,
         context_id: Optional[str] = None,
+        max_buffer_delay_ms: Optional[int] = None,
         duration: Optional[int] = None,
         language: Optional[str] = None,
         stream: bool = True,
@@ -108,6 +109,8 @@ class _TTSContext:
             request_body["add_phoneme_timestamps"] = add_phoneme_timestamps
         if use_original_timestamps:
             request_body["use_original_timestamps"] = use_original_timestamps
+        if max_buffer_delay_ms:
+            request_body["max_buffer_delay_ms"] = max_buffer_delay_ms
 
         if (
             "context_id" in request_body

cartesia/tts/client.py

@@ -5,6 +5,7 @@ from ..core.client_wrapper import SyncClientWrapper
 from .requests.tts_request_voice_specifier import TtsRequestVoiceSpecifierParams
 from .requests.output_format import OutputFormatParams
 from .types.supported_language import SupportedLanguage
+from .types.model_speed import ModelSpeed
 from ..core.request_options import RequestOptions
 from ..core.serialization import convert_and_respect_annotation_metadata
 from json.decoder import JSONDecodeError
@@ -32,13 +33,14 @@ class TtsClient:
         output_format: OutputFormatParams,
         language: typing.Optional[SupportedLanguage] = OMIT,
         duration: typing.Optional[float] = OMIT,
+        speed: typing.Optional[ModelSpeed] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[bytes]:
         """
         Parameters
         ----------
         model_id : str
-            The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+            The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
 
         transcript : str
 
@@ -52,6 +54,8 @@ class TtsClient:
             The maximum duration of the audio in seconds. You do not usually need to specify this.
             If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
 
+        speed : typing.Optional[ModelSpeed]
+
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
 
@@ -92,6 +96,7 @@ class TtsClient:
                     object_=output_format, annotation=OutputFormatParams, direction="write"
                 ),
                 "duration": duration,
+                "speed": speed,
             },
             request_options=request_options,
             omit=OMIT,
@@ -117,13 +122,14 @@ class TtsClient:
         output_format: OutputFormatParams,
         language: typing.Optional[SupportedLanguage] = OMIT,
         duration: typing.Optional[float] = OMIT,
+        speed: typing.Optional[ModelSpeed] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[WebSocketResponse]:
         """
         Parameters
         ----------
         model_id : str
-            The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+            The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
 
         transcript : str
 
@@ -137,6 +143,8 @@ class TtsClient:
             The maximum duration of the audio in seconds. You do not usually need to specify this.
             If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
 
+        speed : typing.Optional[ModelSpeed]
+
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
 
@@ -179,6 +187,7 @@ class TtsClient:
                     object_=output_format, annotation=OutputFormatParams, direction="write"
                 ),
                 "duration": duration,
+                "speed": speed,
             },
             request_options=request_options,
             omit=OMIT,
@@ -218,13 +227,14 @@ class AsyncTtsClient:
         output_format: OutputFormatParams,
         language: typing.Optional[SupportedLanguage] = OMIT,
         duration: typing.Optional[float] = OMIT,
+        speed: typing.Optional[ModelSpeed] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[bytes]:
         """
         Parameters
         ----------
         model_id : str
-            The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+            The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
 
         transcript : str
 
@@ -238,6 +248,8 @@ class AsyncTtsClient:
             The maximum duration of the audio in seconds. You do not usually need to specify this.
             If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
 
+        speed : typing.Optional[ModelSpeed]
+
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration. You can pass in configuration such as `chunk_size`, and more to customize the request and response.
 
@@ -286,6 +298,7 @@ class AsyncTtsClient:
                     object_=output_format, annotation=OutputFormatParams, direction="write"
                 ),
                 "duration": duration,
+                "speed": speed,
             },
             request_options=request_options,
             omit=OMIT,
@@ -311,13 +324,14 @@ class AsyncTtsClient:
         output_format: OutputFormatParams,
         language: typing.Optional[SupportedLanguage] = OMIT,
         duration: typing.Optional[float] = OMIT,
+        speed: typing.Optional[ModelSpeed] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[WebSocketResponse]:
         """
         Parameters
         ----------
         model_id : str
-            The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+            The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
 
         transcript : str
 
@@ -331,6 +345,8 @@ class AsyncTtsClient:
             The maximum duration of the audio in seconds. You do not usually need to specify this.
             If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
 
+        speed : typing.Optional[ModelSpeed]
+
         request_options : typing.Optional[RequestOptions]
             Request-specific configuration.
 
@@ -381,6 +397,7 @@ class AsyncTtsClient:
                     object_=output_format, annotation=OutputFormatParams, direction="write"
                 ),
                 "duration": duration,
+                "speed": speed,
             },
             request_options=request_options,
             omit=OMIT,

cartesia/tts/requests/generation_request.py

@@ -6,6 +6,7 @@ from .tts_request_voice_specifier import TtsRequestVoiceSpecifierParams
 import typing_extensions
 from ..types.supported_language import SupportedLanguage
 from .web_socket_raw_output_format import WebSocketRawOutputFormatParams
+from ..types.model_speed import ModelSpeed
 from ..types.context_id import ContextId
 from ...core.serialization import FieldMetadata
 
@@ -13,7 +14,7 @@ from ...core.serialization import FieldMetadata
 class GenerationRequestParams(typing_extensions.TypedDict):
     model_id: str
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     transcript: typing.Optional[typing.Any]
@@ -30,6 +31,7 @@ class GenerationRequestParams(typing_extensions.TypedDict):
     If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
     """
 
+    speed: typing_extensions.NotRequired[ModelSpeed]
     context_id: typing_extensions.NotRequired[ContextId]
     continue_: typing_extensions.NotRequired[typing_extensions.Annotated[bool, FieldMetadata(alias="continue")]]
     """
@@ -37,6 +39,15 @@ class GenerationRequestParams(typing_extensions.TypedDict):
     If not specified, this defaults to `false`.
     """
 
+    max_buffer_delay_ms: typing_extensions.NotRequired[int]
+    """
+    The maximum time in milliseconds to buffer text before starting generation. Values between [0, 1000]ms are supported. Defaults to 0 (no buffering).
+    
+    When set, the model will buffer incoming text chunks until it's confident it has enough context to generate high-quality speech, or the buffer delay elapses, whichever comes first. Without this option set, the model will kick off generations immediately, ceding control of buffering to the user.
+    
+    Use this to balance responsiveness with higher quality speech generation, which often benefits from having more context.
+    """
+
     flush: typing_extensions.NotRequired[bool]
     """
     Whether to flush the context.

cartesia/tts/requests/tts_request.py

@@ -5,12 +5,13 @@ from .tts_request_voice_specifier import TtsRequestVoiceSpecifierParams
 import typing_extensions
 from ..types.supported_language import SupportedLanguage
 from .output_format import OutputFormatParams
+from ..types.model_speed import ModelSpeed
 
 
 class TtsRequestParams(typing_extensions.TypedDict):
     model_id: str
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     transcript: str
@@ -22,3 +23,5 @@ class TtsRequestParams(typing_extensions.TypedDict):
     The maximum duration of the audio in seconds. You do not usually need to specify this.
     If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
     """
+
+    speed: typing_extensions.NotRequired[ModelSpeed]

cartesia/tts/requests/web_socket_tts_request.py

@@ -5,12 +5,13 @@ import typing_extensions
 from .output_format import OutputFormatParams
 from .tts_request_voice_specifier import TtsRequestVoiceSpecifierParams
 from ...core.serialization import FieldMetadata
+from ..types.model_speed import ModelSpeed
 
 
 class WebSocketTtsRequestParams(typing_extensions.TypedDict):
     model_id: str
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     output_format: typing_extensions.NotRequired[OutputFormatParams]
@@ -23,3 +24,5 @@ class WebSocketTtsRequestParams(typing_extensions.TypedDict):
     add_phoneme_timestamps: typing_extensions.NotRequired[bool]
     continue_: typing_extensions.NotRequired[typing_extensions.Annotated[bool, FieldMetadata(alias="continue")]]
     context_id: typing_extensions.NotRequired[str]
+    max_buffer_delay_ms: typing_extensions.NotRequired[int]
+    speed: typing_extensions.NotRequired[ModelSpeed]

cartesia/tts/types/__init__.py

@@ -6,6 +6,7 @@ from .controls import Controls
 from .emotion import Emotion
 from .flush_id import FlushId
 from .generation_request import GenerationRequest
+from .model_speed import ModelSpeed
 from .mp_3_output_format import Mp3OutputFormat
 from .natural_specifier import NaturalSpecifier
 from .numerical_specifier import NumericalSpecifier
@@ -50,6 +51,7 @@ __all__ = [
     "Emotion",
     "FlushId",
     "GenerationRequest",
+    "ModelSpeed",
     "Mp3OutputFormat",
     "NaturalSpecifier",
     "NumericalSpecifier",

cartesia/tts/types/generation_request.py

@@ -6,6 +6,7 @@ import typing
 from .tts_request_voice_specifier import TtsRequestVoiceSpecifier
 from .supported_language import SupportedLanguage
 from .web_socket_raw_output_format import WebSocketRawOutputFormat
+from .model_speed import ModelSpeed
 from .context_id import ContextId
 import typing_extensions
 from ...core.serialization import FieldMetadata
@@ -15,7 +16,7 @@ from ...core.pydantic_utilities import IS_PYDANTIC_V2
 class GenerationRequest(UniversalBaseModel):
     model_id: str = pydantic.Field()
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     transcript: typing.Optional[typing.Any] = pydantic.Field(default=None)
@@ -32,6 +33,7 @@ class GenerationRequest(UniversalBaseModel):
     If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
     """
 
+    speed: typing.Optional[ModelSpeed] = None
     context_id: typing.Optional[ContextId] = None
     continue_: typing_extensions.Annotated[typing.Optional[bool], FieldMetadata(alias="continue")] = pydantic.Field(
         default=None
@@ -41,6 +43,15 @@ class GenerationRequest(UniversalBaseModel):
     If not specified, this defaults to `false`.
     """
 
+    max_buffer_delay_ms: typing.Optional[int] = pydantic.Field(default=None)
+    """
+    The maximum time in milliseconds to buffer text before starting generation. Values between [0, 1000]ms are supported. Defaults to 0 (no buffering).
+    
+    When set, the model will buffer incoming text chunks until it's confident it has enough context to generate high-quality speech, or the buffer delay elapses, whichever comes first. Without this option set, the model will kick off generations immediately, ceding control of buffering to the user.
+    
+    Use this to balance responsiveness with higher quality speech generation, which often benefits from having more context.
+    """
+
     flush: typing.Optional[bool] = pydantic.Field(default=None)
     """
     Whether to flush the context.

cartesia/tts/types/model_speed.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+ModelSpeed = typing.Union[typing.Literal["slow", "normal", "fast"], typing.Any]

cartesia/tts/types/tts_request.py

@@ -6,13 +6,14 @@ from .tts_request_voice_specifier import TtsRequestVoiceSpecifier
 import typing
 from .supported_language import SupportedLanguage
 from .output_format import OutputFormat
+from .model_speed import ModelSpeed
 from ...core.pydantic_utilities import IS_PYDANTIC_V2
 
 
 class TtsRequest(UniversalBaseModel):
     model_id: str = pydantic.Field()
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     transcript: str
@@ -25,6 +26,8 @@ class TtsRequest(UniversalBaseModel):
     If the duration is not appropriate for the length of the transcript, the output audio may be truncated.
     """
 
+    speed: typing.Optional[ModelSpeed] = None
+
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2
     else:

cartesia/tts/types/web_socket_tts_request.py

@@ -7,13 +7,14 @@ from .output_format import OutputFormat
 from .tts_request_voice_specifier import TtsRequestVoiceSpecifier
 import typing_extensions
 from ...core.serialization import FieldMetadata
+from .model_speed import ModelSpeed
 from ...core.pydantic_utilities import IS_PYDANTIC_V2
 
 
 class WebSocketTtsRequest(UniversalBaseModel):
     model_id: str = pydantic.Field()
     """
-    The ID of the model to use for the generation. See [Models](/build-with-sonic/models) for available models.
+    The ID of the model to use for the generation. See [Models](/build-with-cartesia/models) for available models.
     """
 
     output_format: typing.Optional[OutputFormat] = None
@@ -26,6 +27,8 @@ class WebSocketTtsRequest(UniversalBaseModel):
     add_phoneme_timestamps: typing.Optional[bool] = None
     continue_: typing_extensions.Annotated[typing.Optional[bool], FieldMetadata(alias="continue")] = None
     context_id: typing.Optional[str] = None
+    max_buffer_delay_ms: typing.Optional[int] = None
+    speed: typing.Optional[ModelSpeed] = None
 
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(extra="allow", frozen=True)  # type: ignore # Pydantic v2

cartesia/voices/client.py

@@ -14,8 +14,8 @@ from ..core.api_error import ApiError
 from .. import core
 from ..tts.types.supported_language import SupportedLanguage
 from .types.clone_mode import CloneMode
-from .types.voice_metadata import VoiceMetadata
 from .types.voice_id import VoiceId
+from .types.voice_metadata import VoiceMetadata
 from ..core.jsonable_encoder import jsonable_encoder
 from .types.localize_target_language import LocalizeTargetLanguage
 from .types.gender import Gender
@@ -148,9 +148,9 @@ class VoicesClient:
         name: str,
         language: SupportedLanguage,
         mode: CloneMode,
-        enhance: bool,
         description: typing.Optional[str] = OMIT,
-        transcript: typing.Optional[str] = OMIT,
+        enhance: typing.Optional[bool] = OMIT,
+        base_voice_id: typing.Optional[VoiceId] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> VoiceMetadata:
         """
@@ -177,16 +177,16 @@ class VoicesClient:
             Tradeoff between similarity and stability. Similarity clones sound more like the source clip, but may reproduce background noise. Stability clones always sound like a studio recording, but may not sound as similar to the source clip.
 
 
-        enhance : bool
-            Whether to enhance the clip to improve its quality before cloning. Useful if the clip has background noise.
-
-
         description : typing.Optional[str]
             A description for the voice.
 
 
-        transcript : typing.Optional[str]
-            Optional transcript of the words spoken in the audio clip. Only used for similarity mode.
+        enhance : typing.Optional[bool]
+            Whether to apply AI enhancements to the clip to reduce background noise. This leads to cleaner generated speech at the cost of reduced similarity to the source clip.
+
+
+        base_voice_id : typing.Optional[VoiceId]
+            Optional base voice ID that the cloned voice is derived from.
 
 
         request_options : typing.Optional[RequestOptions]
@@ -208,7 +208,6 @@ class VoicesClient:
             description="Copied from Cartesia docs",
             mode="stability",
             language="en",
-            enhance=True,
         )
         """
         _response = self._client_wrapper.httpx_client.request(
@@ -220,7 +219,7 @@ class VoicesClient:
                 "language": language,
                 "mode": mode,
                 "enhance": enhance,
-                "transcript": transcript,
+                "base_voice_id": base_voice_id,
             },
             files={
                 "clip": clip,
@@ -717,9 +716,9 @@ class AsyncVoicesClient:
         name: str,
         language: SupportedLanguage,
         mode: CloneMode,
-        enhance: bool,
         description: typing.Optional[str] = OMIT,
-        transcript: typing.Optional[str] = OMIT,
+        enhance: typing.Optional[bool] = OMIT,
+        base_voice_id: typing.Optional[VoiceId] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> VoiceMetadata:
         """
@@ -746,16 +745,16 @@ class AsyncVoicesClient:
             Tradeoff between similarity and stability. Similarity clones sound more like the source clip, but may reproduce background noise. Stability clones always sound like a studio recording, but may not sound as similar to the source clip.
 
 
-        enhance : bool
-            Whether to enhance the clip to improve its quality before cloning. Useful if the clip has background noise.
-
-
         description : typing.Optional[str]
             A description for the voice.
 
 
-        transcript : typing.Optional[str]
-            Optional transcript of the words spoken in the audio clip. Only used for similarity mode.
+        enhance : typing.Optional[bool]
+            Whether to apply AI enhancements to the clip to reduce background noise. This leads to cleaner generated speech at the cost of reduced similarity to the source clip.
+
+
+        base_voice_id : typing.Optional[VoiceId]
+            Optional base voice ID that the cloned voice is derived from.
 
 
         request_options : typing.Optional[RequestOptions]
@@ -782,7 +781,6 @@ class AsyncVoicesClient:
                 description="Copied from Cartesia docs",
                 mode="stability",
                 language="en",
-                enhance=True,
             )
 
 
@@ -797,7 +795,7 @@ class AsyncVoicesClient:
                 "language": language,
                 "mode": mode,
                 "enhance": enhance,
-                "transcript": transcript,
+                "base_voice_id": base_voice_id,
             },
             files={
                 "clip": clip,

{cartesia-2.0.0b8.dist-info → cartesia-2.0.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 2.0.0b8
+Version: 2.0.3
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{cartesia-2.0.0b8.dist-info → cartesia-2.0.3.dist-info}/RECORD

@@ -1,15 +1,25 @@
-cartesia/__init__.py,sha256=k-YMKYUtzKObkF9Zn0TuHTC2_Z07mH6CTnZmn1my7po,8143
+cartesia/__init__.py,sha256=p9ilqR-oE6N6pPQbiKUbnfEdDy00ihjNTtg7ES7EHg0,8448
 cartesia/api_status/__init__.py,sha256=_dHNLdknrBjxHtU2PvLumttJM-JTQhJQqhhAQkLqt_U,168
 cartesia/api_status/client.py,sha256=GJ9Dq8iCn3hn8vCIqc6k1fCGEhSz0T0kaPGcdFnbMDY,3146
 cartesia/api_status/requests/__init__.py,sha256=ilEMzEy1JEw484CuL92bX5lHGOznc62pjiDMgiZ0tKM,130
 cartesia/api_status/requests/api_info.py,sha256=AmB6RpquI2yUlTQBtOk8e0qtLmXHYLcGZKpXZahOwmc,172
 cartesia/api_status/types/__init__.py,sha256=6NUyGWiGK1Wl3mXlSMJN2ObKf2LK3vjX2MUP1uopfEQ,118
 cartesia/api_status/types/api_info.py,sha256=o1LwSxnoHpCR7huw9J-cF6LRlC_fiftDQLYUz8p-vTc,568
-cartesia/base_client.py,sha256=EIfMrSkJgMCgzYWJ5GN2RxsWikxcH0kMmcb3WYqfQ_g,6321
+cartesia/auth/__init__.py,sha256=T8_EGgzdzyJLqfD7DAgdkE6G1Ey2sUMyze-7x8HTzGg,355
+cartesia/auth/client.py,sha256=gJurqzV5r-edd5DV2xc1Uy1Fm8Fi3ndaourZmbOh15E,5261
+cartesia/auth/requests/__init__.py,sha256=hR7qCSJCPiOG7f8z8jTKQLOC7QoonSvvPKe0JbcEYEs,278
+cartesia/auth/requests/token_grant.py,sha256=HTrgl6TsokxYIMXeTK-NjSKQ8WQfDwnbEfbyzirs0kk,251
+cartesia/auth/requests/token_request.py,sha256=DeQQhHOLfmL4O3ZqrFq1FtxTDmTsYGpMtmRRiCvxUcE,498
+cartesia/auth/requests/token_response.py,sha256=jXpHZmFe6RWO837e_lC2GJWwqO-b6KHOA-b6tTJVC54,211
+cartesia/auth/types/__init__.py,sha256=iZrkHrlWs8e9KkR27f2IG-B72HC_N05A7Lcyt_EU9SM,242
+cartesia/auth/types/token_grant.py,sha256=sdEqlqS95XSy_Xdp4TEeRSC1hQp4nbPv1HHZFxbU0So,666
+cartesia/auth/types/token_request.py,sha256=2cx2OBXTEjrbuVMOpBzkIm9-DZD2mGiWE6Ui3kumxGI,893
+cartesia/auth/types/token_response.py,sha256=_GcvfQdjwgNu1ODj8EuTkaMsez508a6xuOo8HOVNOJQ,626
+cartesia/base_client.py,sha256=YH0l0UUzanAa9mDdJU6BFQ9XKELiaPTm9NsJpVQ4evA,6539
 cartesia/client.py,sha256=sPAYQLt9W2E_2F17ooocvvJImuNyLrL8xUypgf6dZeI,6238
 cartesia/core/__init__.py,sha256=-t9txgeQZL_1FDw_08GEoj4ft1Cn9Dti6X0Drsadlr0,1519
 cartesia/core/api_error.py,sha256=RE8LELok2QCjABadECTvtDp7qejA1VmINCh6TbqPwSE,426
-cartesia/core/client_wrapper.py,sha256=BEIOireABuSTdCAcsHeQKtZ1D3sIi-CVQv5YFHmfi3Y,1856
+cartesia/core/client_wrapper.py,sha256=xKyHqaGJ7V2OuMGkVOEjxofnt8QaFjgiEIom7EWhBBI,1854
 cartesia/core/datetime_utils.py,sha256=nBys2IsYrhPdszxGKCNRPSOCwa-5DWOHG95FB8G9PKo,1047
 cartesia/core/file.py,sha256=d4NNbX8XvXP32z8KpK2Xovv33nFfruIrpz0QWxlgpZk,2663
 cartesia/core/http_client.py,sha256=KL5RGa0y4n8nX0-07WRg4ZQUTq30sc-XJbWcP5vjBDg,19552
@@ -41,20 +51,20 @@ cartesia/environment.py,sha256=Qnp91BGLic7hXmKsiYub2m3nPfvDWm59aB1wWta1J6A,160
 cartesia/infill/__init__.py,sha256=FTtvy8EDg9nNNg9WCatVgKTRYV8-_v1roeGPAKoa_pw,65
 cartesia/infill/client.py,sha256=PWE5Ak-wsaBM_8g52oDl9PYx76PkW6f900mnxvZf4Bk,12571
 cartesia/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-cartesia/tts/__init__.py,sha256=G0wcYlPrr7hmu5DQgCG7bDTQq36fpP3iBM5164Z0-Js,4701
-cartesia/tts/_async_websocket.py,sha256=U7ySTJqb3V0RDSKPcFfzpBa0pqui05k5BTqiIpSBth0,18652
-cartesia/tts/_websocket.py,sha256=roMJ7oDSjr5U5sTHM8EcGu-EtzbIVUH4HmOY1yI2JL4,19118
-cartesia/tts/client.py,sha256=KMhDaW0gG_uwkSq1EzoC-bCx1G0TLB4K4Gm57L4xDSs,14832
+cartesia/tts/__init__.py,sha256=YrVxJT7i-0wygFgN2hOVftYLEM7JoFgCo3SvLoN7pkA,4735
+cartesia/tts/_async_websocket.py,sha256=tJ-6rdJrviKvGhSW8J8t-rCinXM5gXXQJgDO8OgW3EE,18805
+cartesia/tts/_websocket.py,sha256=Uk6B-TP-0nzeyFE9w-_PzNIiVYP14rKqSDZlm6bU24Q,19271
+cartesia/tts/client.py,sha256=0jWpiNKPp4QbyibfB2tsFb7wqQ9vb4m_QlLB-qxzKP8,15398
 cartesia/tts/requests/__init__.py,sha256=0rcfMLHNbUhkRI1xS09UE4p-WT1BCqrcblFtPxcATOI,3261
 cartesia/tts/requests/cancel_context_request.py,sha256=Wl8g-o5vwl9ENm-H1wsLx441FkIR_4Wt5UYtuWce2Yw,431
 cartesia/tts/requests/controls.py,sha256=xzUJlfgqhaJ1A-JD0LTpoHYk4iEpCuGpSD7qE4YYsRg,285
-cartesia/tts/requests/generation_request.py,sha256=D7eB1HEOzoGbviMavtK1hwsP0lTO7K-pEs3UQIAZpDs,1980
+cartesia/tts/requests/generation_request.py,sha256=cUy--5WFsC7E-KEPVay3QpU_-a3GVGnatTKsBw6hIV4,2743
 cartesia/tts/requests/mp_3_output_format.py,sha256=PGDVzC1d7-Jce12rFxtF8G1pTHmlUdiGAhykFTABg0w,316
 cartesia/tts/requests/output_format.py,sha256=8TKu9AAeHCR5L4edzYch8FIYIldn4bM7ySrsCl8W_g8,842
 cartesia/tts/requests/phoneme_timestamps.py,sha256=ft81nmqElZAnvTBT27lY6YWfF18ZGsCx3Y1XHv9J7cM,267
 cartesia/tts/requests/raw_output_format.py,sha256=S60Vp7DeAATCMLF3bXgxhw0zILJBWJ9GhI9irAg_UkI,316
 cartesia/tts/requests/speed.py,sha256=-YGBWwh7_VtCBnYlT5EVsnrmcHFMEBTxy9LathZhkMA,259
-cartesia/tts/requests/tts_request.py,sha256=rh7akZLf_w0RukKclCic9fKIIJi-X1M1GeHnJ14rjKk,921
+cartesia/tts/requests/tts_request.py,sha256=KBoahYfPbDENlEWsqnR4z1ZIhGIJwhLrzQIzkbtqtzE,1021
 cartesia/tts/requests/tts_request_embedding_specifier.py,sha256=-M54ZjV0H5LPwcKtz0bOVqlkvO1pPiMbqMbVBMko3Ns,565
 cartesia/tts/requests/tts_request_id_specifier.py,sha256=-0ClfyJnnaH0uAcF5r84s3cM_cw2wT39dp6T4JYzOQ8,536
 cartesia/tts/requests/tts_request_voice_specifier.py,sha256=eGzL4aVGq4gKPxeglsV7-wuhxg8x33Qth3uFTTytgeI,337
@@ -71,16 +81,17 @@ cartesia/tts/requests/web_socket_response.py,sha256=WqZ6RgO4suG78wiVSIsOWwyXBioV
 cartesia/tts/requests/web_socket_stream_options.py,sha256=VIvblFw9hGZvDzFpOnC11G0NvrFSVt-1-0sY5rpcZPI,232
 cartesia/tts/requests/web_socket_timestamps_response.py,sha256=MK3zN2Q_PVWJtX5DidNB0uXoF2o33rv6qCYPVaourxY,351
 cartesia/tts/requests/web_socket_tts_output.py,sha256=pX2uf0XVdziFhXCydwLlVOWb-LvBiuq-cBI6R1INiMg,760
-cartesia/tts/requests/web_socket_tts_request.py,sha256=jv8EYSxsjb063uzKjybRRNusmNtfzt516-r2urfG-vU,1101
+cartesia/tts/requests/web_socket_tts_request.py,sha256=RBFrdmYe0SRlzhEUwhTFCL8ZC1tbIwD2aFnUgF-my80,1260
 cartesia/tts/requests/word_timestamps.py,sha256=WMfBJtETi6wTpES0pYZCFfFRfEbzWE-RtosDJ5seUWg,261
 cartesia/tts/socket_client.py,sha256=zTPayHbgy-yQQ50AE1HXN4GMyanisZcLXf7Ds1paYks,11621
-cartesia/tts/types/__init__.py,sha256=yV_-DY9EPNAFEfuIk3wgRLcc4Ta5igv0T5g-IIQ53v0,3251
+cartesia/tts/types/__init__.py,sha256=44KWnBiqkNHZZpy8M5uze8cdEFE79sbMRVJStxQSEhM,3305
 cartesia/tts/types/cancel_context_request.py,sha256=zInhk3qRZsSc0F1aYJ-Q5BHJsosTrb22IJWhzue-eKE,856
 cartesia/tts/types/context_id.py,sha256=UCEtq5xFGOeBCECcY6Y-gYVe_Peg1hFhH9YYOkpApQg,81
 cartesia/tts/types/controls.py,sha256=H4CSu79mM1Ld4NZx_5uXw3EwRzTEMQRxKBRvFpcFb8Y,644
 cartesia/tts/types/emotion.py,sha256=zocyDcHTiFFnNRgo2YLMi70iGyffa080B4mkg9lcqVc,764
 cartesia/tts/types/flush_id.py,sha256=HCIKo9o8d7YWKtaSNU3TEvfUVBju93ckGQy01Z9wLcE,79
-cartesia/tts/types/generation_request.py,sha256=Ig14IIulKgsHtIoRmfrov-U0jpYWQqOjCAUt5Fk20z4,2476
+cartesia/tts/types/generation_request.py,sha256=H5ZaNGH2ngTZY-NQ7wqLUiIAArH8KFo3rt690bxWCUw,3242
+cartesia/tts/types/model_speed.py,sha256=iiTj8V0piFCX2FZh5B8EkgRhZDlj4z3VFcQhp66e7y8,160
 cartesia/tts/types/mp_3_output_format.py,sha256=0WGblkuDUL7pZO1aRuQ_mU2Z5gN9xIabRfRKkjtzms8,731
 cartesia/tts/types/natural_specifier.py,sha256=K526P1RRuBGy80hyd_tX8tohPrE8DR9EgTCxS5wce0o,188
 cartesia/tts/types/numerical_specifier.py,sha256=tJpIskWO545luCKMFM9JlVc7VVhBhSvqL1qurhzL9cI,92
@@ -90,7 +101,7 @@ cartesia/tts/types/raw_encoding.py,sha256=eyc2goiYOTxWcuKHAgYZ2SrnfePW22Fbmc-5fG
 cartesia/tts/types/raw_output_format.py,sha256=jZGVaS0KIi9mU6trfskgA3HbMKJolhrwICnuDhF01ic,673
 cartesia/tts/types/speed.py,sha256=4c5WdxocBw6WSMnundSaNnceUeooU0vikhy00FW6M-w,239
 cartesia/tts/types/supported_language.py,sha256=riDRduThMbMWAq9i2uCfxhwVTpgaFwNDZ9LhEIl4zHY,237
-cartesia/tts/types/tts_request.py,sha256=MxWcMLxIpotkPiPhFIhdHTtDzYv8yzLJwrwX3kBhkIg,1290
+cartesia/tts/types/tts_request.py,sha256=FGcxW-siiQpEzJZSHMET3nDSYHSzRt3WSTO-cCEz9u4,1376
 cartesia/tts/types/tts_request_embedding_specifier.py,sha256=eL_qCEr4pvWfy4qp9hZBuVdCincX5DBVqfv1vLt2_Vk,942
 cartesia/tts/types/tts_request_id_specifier.py,sha256=ktGdkkTRQ9scA-lt8qJ2jn_E5WzoOK8AXMrVqi71gf0,906
 cartesia/tts/types/tts_request_voice_specifier.py,sha256=p-3UQ62uFL1SgbX73Ex1D_V73Ef0wmT1ApOt1iLZmwE,307
@@ -107,7 +118,7 @@ cartesia/tts/types/web_socket_response.py,sha256=mHDECZ4K84QmN2s0IWuBsXBt83Yq7Qx
 cartesia/tts/types/web_socket_stream_options.py,sha256=MhDSxBFqMuQeWjoyPqXVnTEzLjF8g6aojeigb5dQUgU,596
 cartesia/tts/types/web_socket_timestamps_response.py,sha256=kuWXI82ncF1QapnaHEjwrL84qWob7ByQU-yh1e0IEmk,667
 cartesia/tts/types/web_socket_tts_output.py,sha256=uvkv0smTBhdm18Rl17C0Ml4Inh79YBHNzAcKnZBs14Y,979
-cartesia/tts/types/web_socket_tts_request.py,sha256=4qp-mPmVZOMlHAr7f8ABMWYS3cy5OWPjxDNeWayU0aE,1429
+cartesia/tts/types/web_socket_tts_request.py,sha256=TlqUQPhqZcDZ6jgpzLqzJnYVGBIc9_TQYSu_SVYkVVk,1567
 cartesia/tts/types/word_timestamps.py,sha256=XZ2Q0prdb3F9c3AiOKXu4s3A3jBxE-qIt1npHOf16R0,631
 cartesia/tts/utils/constants.py,sha256=1CHa5flJf8--L_eYyOyOiWJNZ-Q81ufHZxDbJs8xYSk,418
 cartesia/tts/utils/tts.py,sha256=u7PgPxlJs6fcQTfr-jqAvBCAaK3JWLhF5QF4s-PwoMo,2093
@@ -121,7 +132,7 @@ cartesia/voice_changer/types/__init__.py,sha256=qAiHsdRpnFeS0lBkYp_NRrhSJiRXCg5-
 cartesia/voice_changer/types/output_format_container.py,sha256=RqLDELdgeOjYqNTJX1Le62qjiFiJGxf0cYnol88-LLM,166
 cartesia/voice_changer/types/streaming_response.py,sha256=rQ4ZehtOHsCBKijyULz_ahGQYNj1yus6AM6u2wgcBsI,1963
 cartesia/voices/__init__.py,sha256=2D58Bir45LvcvP08QMnPlFE8DD8BONTjPLkIDdKs7vg,1891
-cartesia/voices/client.py,sha256=8zQZAtaCAJi79puMxVhzR5OWCDjows53k4oTvSgcdJM,38867
+cartesia/voices/client.py,sha256=A_PEoCLko1znexKKicp-gZVMUcSpDoKqz3p1r4Aa04k,38993
 cartesia/voices/requests/__init__.py,sha256=XiBJbSYeQCgFMtwywKvQ0Nmp7Zf_0WskzRhgr9c8h38,1072
 cartesia/voices/requests/create_voice_request.py,sha256=r6dKb9ga0ZsAi_6PXuE43u2lLgfQg2DIYjk2Neng7pI,617
 cartesia/voices/requests/embedding_response.py,sha256=PGZkBD8UBcv2MYQbBXyD4T6lzaE9oSGGwXx-MoXCp0M,228
@@ -160,6 +171,6 @@ cartesia/voices/types/voice_expand_options.py,sha256=e4FroWdlxEE-LXQfT1RWlGHtswl
 cartesia/voices/types/voice_id.py,sha256=GDoXcRVeIm-V21R4suxG2zqLD3DLYkXE9kgizadzFKo,79
 cartesia/voices/types/voice_metadata.py,sha256=4KNGjXMUKm3niv-NvKIFVGtiilpH13heuzKcZYNQxk4,1181
 cartesia/voices/types/weight.py,sha256=XqDU7_JItNUb5QykIDqTbELlRYQdbt2SviRgW0w2LKo,80
-cartesia-2.0.0b8.dist-info/METADATA,sha256=ynQsxGb1v5ZHMnXkeqYceRFrC-bxwuRaopOPyuBbCsk,11208
-cartesia-2.0.0b8.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
-cartesia-2.0.0b8.dist-info/RECORD,,
+cartesia-2.0.3.dist-info/METADATA,sha256=cW9xivCIN1lB-8xc8V_-DMwh0pwJa3gwmsYO3XwJl0M,11206
+cartesia-2.0.3.dist-info/WHEEL,sha256=Zb28QaM1gQi8f4VCBhsUklF61CTlNYfs9YAZn-TOGFk,88
+cartesia-2.0.3.dist-info/RECORD,,