cartesia 2.0.0a0__py3-none-any.whl → 2.0.0b1__py3-none-any.whl

cartesia/__init__.py

@@ -126,6 +126,8 @@ from .voices import (
     LocalizeDialect,
     LocalizeDialectParams,
     LocalizeEnglishDialect,
+    LocalizePortugueseDialect,
+    LocalizeSpanishDialect,
     LocalizeTargetLanguage,
     LocalizeVoiceRequest,
     LocalizeVoiceRequestParams,
@@ -180,6 +182,8 @@ __all__ = [
     "LocalizeDialect",
     "LocalizeDialectParams",
     "LocalizeEnglishDialect",
+    "LocalizePortugueseDialect",
+    "LocalizeSpanishDialect",
     "LocalizeTargetLanguage",
     "LocalizeVoiceRequest",
     "LocalizeVoiceRequestParams",

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": "2.0.0a0",
+            "X-Fern-SDK-Version": "2.0.0b1",
         }
         headers["X-API-Key"] = self.api_key
         headers["Cartesia-Version"] = "2024-06-10"

cartesia/core/http_client.py

@@ -85,8 +85,8 @@ def _retry_timeout(response: httpx.Response, retries: int) -> float:
 
 
 def _should_retry(response: httpx.Response) -> bool:
-    retriable_400s = [429, 408, 409]
-    return response.status_code >= 500 or response.status_code in retriable_400s
+    retryable_400s = [429, 408, 409]
+    return response.status_code >= 500 or response.status_code in retryable_400s
 
 
 def remove_omit_from_dict(

cartesia/core/pydantic_utilities.py

@@ -79,7 +79,7 @@ def to_jsonable_with_fallback(
 class UniversalBaseModel(pydantic.BaseModel):
     if IS_PYDANTIC_V2:
         model_config: typing.ClassVar[pydantic.ConfigDict] = pydantic.ConfigDict(
-            # Allow fields begining with `model_` to be used in the model
+            # Allow fields beginning with `model_` to be used in the model
             protected_namespaces=(),
         )  # type: ignore # Pydantic v2
 
@@ -128,7 +128,7 @@ class UniversalBaseModel(pydantic.BaseModel):
         Override the default dict method to `exclude_unset` by default. This function patches
         `exclude_unset` to work include fields within non-None default values.
         """
-        # Note: the logic here is multi-plexed given the levers exposed in Pydantic V1 vs V2
+        # Note: the logic here is multiplexed given the levers exposed in Pydantic V1 vs V2
         # Pydantic V1's .dict can be extremely slow, so we do not want to call it twice.
         #
         # We'd ideally do the same for Pydantic V2, but it shells out to a library to serialize models

cartesia/datasets/client.py

@@ -82,7 +82,7 @@ class DatasetsClient:
             api_key="YOUR_API_KEY",
         )
         client.datasets.create(
-            name="string",
+            name="name",
         )
         """
         _response = self._client_wrapper.httpx_client.request(
@@ -129,7 +129,7 @@ class DatasetsClient:
             api_key="YOUR_API_KEY",
         )
         client.datasets.list_files(
-            id="string",
+            id="id",
         )
         """
         _response = self._client_wrapper.httpx_client.request(
@@ -170,17 +170,6 @@ class DatasetsClient:
         Returns
         -------
         None
-
-        Examples
-        --------
-        from cartesia import Cartesia
-
-        client = Cartesia(
-            api_key="YOUR_API_KEY",
-        )
-        client.datasets.upload_file(
-            id="string",
-        )
         """
         _response = self._client_wrapper.httpx_client.request(
             f"datasets/{jsonable_encoder(id)}/files",
@@ -280,7 +269,7 @@ class AsyncDatasetsClient:
 
         async def main() -> None:
             await client.datasets.create(
-                name="string",
+                name="name",
             )
 
 
@@ -337,7 +326,7 @@ class AsyncDatasetsClient:
 
         async def main() -> None:
             await client.datasets.list_files(
-                id="string",
+                id="id",
             )
 
 
@@ -381,25 +370,6 @@ class AsyncDatasetsClient:
         Returns
         -------
         None
-
-        Examples
-        --------
-        import asyncio
-
-        from cartesia import AsyncCartesia
-
-        client = AsyncCartesia(
-            api_key="YOUR_API_KEY",
-        )
-
-
-        async def main() -> None:
-            await client.datasets.upload_file(
-                id="string",
-            )
-
-
-        asyncio.run(main())
         """
         _response = await self._client_wrapper.httpx_client.request(
             f"datasets/{jsonable_encoder(id)}/files",

cartesia/infill/client.py

@@ -34,16 +34,26 @@ class InfillClient:
         output_format_encoding: typing.Optional[RawEncoding] = OMIT,
         output_format_bit_rate: typing.Optional[int] = OMIT,
         voice_experimental_controls_speed: typing.Optional[Speed] = OMIT,
-        voice_experimental_controls_emotion: typing.Optional[Emotion] = OMIT,
+        voice_experimental_controls_emotion: typing.Optional[typing.List[Emotion]] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[bytes]:
         """
         Generate audio that smoothly connects two existing audio segments. This is useful for inserting new speech between existing speech segments while maintaining natural transitions.
 
+        **The cost is 1 credit per character of the infill text plus a fixed cost of 300 credits.**
+
         Only the `sonic-preview` model is supported for infill at this time.
 
         At least one of `left_audio` or `right_audio` must be provided.
 
+        As with all generative models, there's some inherent variability, but here's some tips we recommend to get the best results from infill:
+        - Use longer infill transcripts
+          - This gives the model more flexibility to adapt to the rest of the audio
+        - Target natural pauses in the audio when deciding where to clip
+          - This means you don't need word-level timestamps to be as precise
+        - Clip right up to the start and end of the audio segment you want infilled, keeping as much silence in the left/right audio segments as possible
+          - This helps the model generate more natural transitions
+
         Parameters
         ----------
         left_audio : core.File
@@ -84,7 +94,7 @@ class InfillClient:
             If you specify a number, 0.0 is the default speed, -1.0 is the slowest speed, and 1.0 is the fastest speed.
 
 
-        voice_experimental_controls_emotion : typing.Optional[Emotion]
+        voice_experimental_controls_emotion : typing.Optional[typing.List[Emotion]]
             An array of emotion:level tags.
 
             Supported emotions are: anger, positivity, surprise, sadness, and curiosity.
@@ -114,16 +124,18 @@ class InfillClient:
             output_format_container="mp3",
             output_format_sample_rate=44100,
             output_format_bit_rate=128000,
+            voice_experimental_controls_speed="slowest",
+            voice_experimental_controls_emotion=["surprise:high", "curiosity:high"],
         )
         """
         with self._client_wrapper.httpx_client.stream(
             "infill/bytes",
             method="POST",
             data={
-                "model_id[]": model_id,
-                "language[]": language,
-                "transcript[]": transcript,
-                "voice[id]": voice_id,
+                "model_id": model_id,
+                "language": language,
+                "transcript": transcript,
+                "voice_id": voice_id,
                 "output_format[container]": output_format_container,
                 "output_format[sample_rate]": output_format_sample_rate,
                 "output_format[encoding]": output_format_encoding,
@@ -169,16 +181,26 @@ class AsyncInfillClient:
         output_format_encoding: typing.Optional[RawEncoding] = OMIT,
         output_format_bit_rate: typing.Optional[int] = OMIT,
         voice_experimental_controls_speed: typing.Optional[Speed] = OMIT,
-        voice_experimental_controls_emotion: typing.Optional[Emotion] = OMIT,
+        voice_experimental_controls_emotion: typing.Optional[typing.List[Emotion]] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[bytes]:
         """
         Generate audio that smoothly connects two existing audio segments. This is useful for inserting new speech between existing speech segments while maintaining natural transitions.
 
+        **The cost is 1 credit per character of the infill text plus a fixed cost of 300 credits.**
+
         Only the `sonic-preview` model is supported for infill at this time.
 
         At least one of `left_audio` or `right_audio` must be provided.
 
+        As with all generative models, there's some inherent variability, but here's some tips we recommend to get the best results from infill:
+        - Use longer infill transcripts
+          - This gives the model more flexibility to adapt to the rest of the audio
+        - Target natural pauses in the audio when deciding where to clip
+          - This means you don't need word-level timestamps to be as precise
+        - Clip right up to the start and end of the audio segment you want infilled, keeping as much silence in the left/right audio segments as possible
+          - This helps the model generate more natural transitions
+
         Parameters
         ----------
         left_audio : core.File
@@ -219,7 +241,7 @@ class AsyncInfillClient:
             If you specify a number, 0.0 is the default speed, -1.0 is the slowest speed, and 1.0 is the fastest speed.
 
 
-        voice_experimental_controls_emotion : typing.Optional[Emotion]
+        voice_experimental_controls_emotion : typing.Optional[typing.List[Emotion]]
             An array of emotion:level tags.
 
             Supported emotions are: anger, positivity, surprise, sadness, and curiosity.
@@ -254,6 +276,8 @@ class AsyncInfillClient:
                 output_format_container="mp3",
                 output_format_sample_rate=44100,
                 output_format_bit_rate=128000,
+                voice_experimental_controls_speed="slowest",
+                voice_experimental_controls_emotion=["surprise:high", "curiosity:high"],
             )
 
 
@@ -263,10 +287,10 @@ class AsyncInfillClient:
             "infill/bytes",
             method="POST",
             data={
-                "model_id[]": model_id,
-                "language[]": language,
-                "transcript[]": transcript,
-                "voice[id]": voice_id,
+                "model_id": model_id,
+                "language": language,
+                "transcript": transcript,
+                "voice_id": voice_id,
                 "output_format[container]": output_format_container,
                 "output_format[sample_rate]": output_format_sample_rate,
                 "output_format[encoding]": output_format_encoding,

cartesia/tts/client.py

@@ -67,7 +67,7 @@ class TtsClient:
             api_key="YOUR_API_KEY",
         )
         client.tts.bytes(
-            model_id="sonic-english",
+            model_id="sonic",
             transcript="Hello, world!",
             voice={"mode": "id", "id": "694f9389-aac1-45b6-b726-9d9369183238"},
             language="en",
@@ -152,7 +152,7 @@ class TtsClient:
             api_key="YOUR_API_KEY",
         )
         response = client.tts.sse(
-            model_id="sonic-english",
+            model_id="sonic",
             transcript="Hello, world!",
             voice={"mode": "id", "id": "694f9389-aac1-45b6-b726-9d9369183238"},
             language="en",
@@ -258,7 +258,7 @@ class AsyncTtsClient:
 
         async def main() -> None:
             await client.tts.bytes(
-                model_id="sonic-english",
+                model_id="sonic",
                 transcript="Hello, world!",
                 voice={"mode": "id", "id": "694f9389-aac1-45b6-b726-9d9369183238"},
                 language="en",
@@ -351,7 +351,7 @@ class AsyncTtsClient:
 
         async def main() -> None:
             response = await client.tts.sse(
-                model_id="sonic-english",
+                model_id="sonic",
                 transcript="Hello, world!",
                 voice={"mode": "id", "id": "694f9389-aac1-45b6-b726-9d9369183238"},
                 language="en",

cartesia/voices/__init__.py

@@ -10,6 +10,8 @@ from .types import (
     IdSpecifier,
     LocalizeDialect,
     LocalizeEnglishDialect,
+    LocalizePortugueseDialect,
+    LocalizeSpanishDialect,
     LocalizeTargetLanguage,
     LocalizeVoiceRequest,
     MixVoiceSpecifier,
@@ -49,6 +51,8 @@ __all__ = [
     "LocalizeDialect",
     "LocalizeDialectParams",
     "LocalizeEnglishDialect",
+    "LocalizePortugueseDialect",
+    "LocalizeSpanishDialect",
     "LocalizeTargetLanguage",
     "LocalizeVoiceRequest",
     "LocalizeVoiceRequestParams",