cartesia 2.0.0b7__tar.gz → 2.0.0b8__tar.gz

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 2.0.0b7
+Version: 2.0.0b8
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers
@@ -47,53 +47,6 @@ Our complete API documentation can be found [on docs.cartesia.ai](https://docs.c
 pip install cartesia
 ```
 
-## Reference
-
-A full reference for this library is available [here](./reference.md).
-
-## Voices
-
-```python
-from cartesia import Cartesia
-import os
-
-client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
-
-# Get all available voices
-voices = client.voices.list()
-print(voices)
-
-# Get a specific voice
-voice = client.voices.get(id="a0e99841-438c-4a64-b679-ae501e7d6091")
-print("The embedding for", voice.name, "is", voice.embedding)
-
-# Clone a voice using file data
-cloned_voice = client.voices.clone(
-    clip=open("path/to/voice.wav", "rb"),
-    name="Test cloned voice",
-    language="en",
-    mode="similarity",  # or "stability"
-    enhance=False, # use enhance=True to clean and denoise the cloning audio
-    description="Test voice description"
-)
-
-# Mix voices together
-mixed_voice = client.voices.mix(
-    voices=[
-        {"id": "voice_id_1", "weight": 0.25},
-        {"id": "voice_id_2", "weight": 0.75}
-    ]
-)
-
-# Create a new voice from embedding
-new_voice = client.voices.create(
-    name="Test Voice",
-    description="Test voice description",
-    embedding=[...],  # List[float] with 192 dimensions
-    language="en"
-)
-```
-
 ## Usage
 
 Instantiate and use the client with the following:
@@ -112,10 +65,6 @@ client.tts.bytes(
     voice={
         "mode": "id",
         "id": "694f9389-aac1-45b6-b726-9d9369183238",
-        "experimental_controls": {
-            "speed": 0.5,  # range between [-1.0, 1.0], or "slow", "fastest", etc.
-            "emotion": ["positivity", "curiosity:low"] # list of emotions with optional intensity
-        }
     },
     language="en",
     output_format={
@@ -176,7 +125,7 @@ except ApiError as e:
 
 ## Streaming
 
-The SDK supports streaming responses, as well, the response will be a generator that you can loop over.
+The SDK supports streaming responses as well, returning a generator that you can iterate over with a `for ... in ...` loop:
 
 ```python
 from cartesia import Cartesia
@@ -215,7 +164,9 @@ for chunk in chunks:
     print(f"Received chunk of size: {len(chunk.data)}")
 ```
 
-## WebSocket
+## WebSockets
+
+For the lowest latency in advanced usecases (such as streaming in an LLM-generated transcript and streaming out audio), you should use our websockets client:
 
 ```python
 from cartesia import Cartesia
@@ -223,15 +174,10 @@ from cartesia.tts import TtsRequestEmbeddingSpecifierParams, OutputFormat_RawPar
 import pyaudio
 import os
 
-client = Cartesia(
-    api_key=os.getenv("CARTESIA_API_KEY"),
-)
+client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
 voice_id = "a0e99841-438c-4a64-b679-ae501e7d6091"
 transcript = "Hello! Welcome to Cartesia"
 
-# You can check out our models at https://docs.cartesia.ai/getting-started/available-models
-model_id = "sonic-2"
-
 p = pyaudio.PyAudio()
 rate = 22050
 
@@ -242,14 +188,14 @@ ws = client.tts.websocket()
 
 # Generate and stream audio using the websocket
 for output in ws.send(
-    model_id=model_id,
+    model_id="sonic-2", # see: https://docs.cartesia.ai/getting-started/available-models
     transcript=transcript,
     voice={"id": voice_id},
     stream=True,
     output_format={
         "container": "raw",
         "encoding": "pcm_f32le",
-        "sample_rate": 22050
+        "sample_rate": rate
     },
 ):
     buffer = output.audio
@@ -267,6 +213,40 @@ p.terminate()
 ws.close()  # Close the websocket connection
 ```
 
+## Voices
+
+List all available Voices with `client.voices.list`, which returns an iterable that automatically handles pagination:
+
+```python
+from cartesia import Cartesia
+import os
+
+client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
+
+# Get all available Voices
+voices = client.voices.list()
+for voice in voices:
+    print(voice)
+```
+
+You can also get the complete metadata for a specific Voice, or make a new Voice by cloning from an audio sample:
+
+```python
+# Get a specific Voice
+voice = client.voices.get(id="a0e99841-438c-4a64-b679-ae501e7d6091")
+print("The embedding for", voice.name, "is", voice.embedding)
+
+# Clone a Voice using file data
+cloned_voice = client.voices.clone(
+    clip=open("path/to/voice.wav", "rb"),
+    name="Test cloned voice",
+    language="en",
+    mode="similarity",  # or "stability"
+    enhance=False, # use enhance=True to clean and denoise the cloning audio
+    description="Test voice description"
+)
+```
+
 ## Requesting Timestamps
 
 ```python
@@ -290,7 +270,8 @@ async def main():
             "encoding": "pcm_f32le",
             "sample_rate": 44100
         },
-        add_timestamps=True,  # Enable word-level timestamps
+        add_timestamps=True,            # Enable word-level timestamps
+        add_phoneme_timestamps=True,    # Enable phonemized timestamps
         stream=True
     )
 
@@ -358,6 +339,26 @@ client.tts.bytes(..., request_options={
 })
 ```
 
+### Mixing voices and creating from embeddings
+
+```python
+# Mix voices together
+mixed_voice = client.voices.mix(
+    voices=[
+        {"id": "voice_id_1", "weight": 0.25},
+        {"id": "voice_id_2", "weight": 0.75}
+    ]
+)
+
+# Create a new voice from embedding
+new_voice = client.voices.create(
+    name="Test Voice",
+    description="Test voice description",
+    embedding=[...],  # List[float] with 192 dimensions
+    language="en"
+)
+```
+
 ### Custom Client
 
 You can override the `httpx` client to customize it for your use-case. Some common use-cases include support for proxies
@@ -375,6 +376,10 @@ client = Cartesia(
 )
 ```
 
+## Reference
+
+A full reference for this library is available [here](./reference.md).
+
 ## Contributing
 
 Note that most of this library is generated programmatically from

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/README.md

@@ -15,53 +15,6 @@ Our complete API documentation can be found [on docs.cartesia.ai](https://docs.c
 pip install cartesia
 ```
 
-## Reference
-
-A full reference for this library is available [here](./reference.md).
-
-## Voices
-
-```python
-from cartesia import Cartesia
-import os
-
-client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
-
-# Get all available voices
-voices = client.voices.list()
-print(voices)
-
-# Get a specific voice
-voice = client.voices.get(id="a0e99841-438c-4a64-b679-ae501e7d6091")
-print("The embedding for", voice.name, "is", voice.embedding)
-
-# Clone a voice using file data
-cloned_voice = client.voices.clone(
-    clip=open("path/to/voice.wav", "rb"),
-    name="Test cloned voice",
-    language="en",
-    mode="similarity",  # or "stability"
-    enhance=False, # use enhance=True to clean and denoise the cloning audio
-    description="Test voice description"
-)
-
-# Mix voices together
-mixed_voice = client.voices.mix(
-    voices=[
-        {"id": "voice_id_1", "weight": 0.25},
-        {"id": "voice_id_2", "weight": 0.75}
-    ]
-)
-
-# Create a new voice from embedding
-new_voice = client.voices.create(
-    name="Test Voice",
-    description="Test voice description",
-    embedding=[...],  # List[float] with 192 dimensions
-    language="en"
-)
-```
-
 ## Usage
 
 Instantiate and use the client with the following:
@@ -80,10 +33,6 @@ client.tts.bytes(
     voice={
         "mode": "id",
         "id": "694f9389-aac1-45b6-b726-9d9369183238",
-        "experimental_controls": {
-            "speed": 0.5,  # range between [-1.0, 1.0], or "slow", "fastest", etc.
-            "emotion": ["positivity", "curiosity:low"] # list of emotions with optional intensity
-        }
     },
     language="en",
     output_format={
@@ -144,7 +93,7 @@ except ApiError as e:
 
 ## Streaming
 
-The SDK supports streaming responses, as well, the response will be a generator that you can loop over.
+The SDK supports streaming responses as well, returning a generator that you can iterate over with a `for ... in ...` loop:
 
 ```python
 from cartesia import Cartesia
@@ -183,7 +132,9 @@ for chunk in chunks:
     print(f"Received chunk of size: {len(chunk.data)}")
 ```
 
-## WebSocket
+## WebSockets
+
+For the lowest latency in advanced usecases (such as streaming in an LLM-generated transcript and streaming out audio), you should use our websockets client:
 
 ```python
 from cartesia import Cartesia
@@ -191,15 +142,10 @@ from cartesia.tts import TtsRequestEmbeddingSpecifierParams, OutputFormat_RawPar
 import pyaudio
 import os
 
-client = Cartesia(
-    api_key=os.getenv("CARTESIA_API_KEY"),
-)
+client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
 voice_id = "a0e99841-438c-4a64-b679-ae501e7d6091"
 transcript = "Hello! Welcome to Cartesia"
 
-# You can check out our models at https://docs.cartesia.ai/getting-started/available-models
-model_id = "sonic-2"
-
 p = pyaudio.PyAudio()
 rate = 22050
 
@@ -210,14 +156,14 @@ ws = client.tts.websocket()
 
 # Generate and stream audio using the websocket
 for output in ws.send(
-    model_id=model_id,
+    model_id="sonic-2", # see: https://docs.cartesia.ai/getting-started/available-models
     transcript=transcript,
     voice={"id": voice_id},
     stream=True,
     output_format={
         "container": "raw",
         "encoding": "pcm_f32le",
-        "sample_rate": 22050
+        "sample_rate": rate
     },
 ):
     buffer = output.audio
@@ -235,6 +181,40 @@ p.terminate()
 ws.close()  # Close the websocket connection
 ```
 
+## Voices
+
+List all available Voices with `client.voices.list`, which returns an iterable that automatically handles pagination:
+
+```python
+from cartesia import Cartesia
+import os
+
+client = Cartesia(api_key=os.getenv("CARTESIA_API_KEY"))
+
+# Get all available Voices
+voices = client.voices.list()
+for voice in voices:
+    print(voice)
+```
+
+You can also get the complete metadata for a specific Voice, or make a new Voice by cloning from an audio sample:
+
+```python
+# Get a specific Voice
+voice = client.voices.get(id="a0e99841-438c-4a64-b679-ae501e7d6091")
+print("The embedding for", voice.name, "is", voice.embedding)
+
+# Clone a Voice using file data
+cloned_voice = client.voices.clone(
+    clip=open("path/to/voice.wav", "rb"),
+    name="Test cloned voice",
+    language="en",
+    mode="similarity",  # or "stability"
+    enhance=False, # use enhance=True to clean and denoise the cloning audio
+    description="Test voice description"
+)
+```
+
 ## Requesting Timestamps
 
 ```python
@@ -258,7 +238,8 @@ async def main():
             "encoding": "pcm_f32le",
             "sample_rate": 44100
         },
-        add_timestamps=True,  # Enable word-level timestamps
+        add_timestamps=True,            # Enable word-level timestamps
+        add_phoneme_timestamps=True,    # Enable phonemized timestamps
         stream=True
     )
 
@@ -326,6 +307,26 @@ client.tts.bytes(..., request_options={
 })
 ```
 
+### Mixing voices and creating from embeddings
+
+```python
+# Mix voices together
+mixed_voice = client.voices.mix(
+    voices=[
+        {"id": "voice_id_1", "weight": 0.25},
+        {"id": "voice_id_2", "weight": 0.75}
+    ]
+)
+
+# Create a new voice from embedding
+new_voice = client.voices.create(
+    name="Test Voice",
+    description="Test voice description",
+    embedding=[...],  # List[float] with 192 dimensions
+    language="en"
+)
+```
+
 ### Custom Client
 
 You can override the `httpx` client to customize it for your use-case. Some common use-cases include support for proxies
@@ -343,6 +344,10 @@ client = Cartesia(
 )
 ```
 
+## Reference
+
+A full reference for this library is available [here](./reference.md).
+
 ## Contributing
 
 Note that most of this library is generated programmatically from

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/pyproject.toml

@@ -3,7 +3,7 @@ name = "cartesia"
 
 [tool.poetry]
 name = "cartesia"
-version = "2.0.0b7"
+version = "2.0.0b8"
 description = ""
 readme = "README.md"
 authors = []

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/src/cartesia/__init__.py

@@ -129,6 +129,9 @@ from .voices import (
     LocalizeDialect,
     LocalizeDialectParams,
     LocalizeEnglishDialect,
+    LocalizeFrenchDialect,
+    LocalizePortugueseDialect,
+    LocalizeSpanishDialect,
     LocalizeTargetLanguage,
     LocalizeVoiceRequest,
     LocalizeVoiceRequestParams,
@@ -187,6 +190,9 @@ __all__ = [
     "LocalizeDialect",
     "LocalizeDialectParams",
     "LocalizeEnglishDialect",
+    "LocalizeFrenchDialect",
+    "LocalizePortugueseDialect",
+    "LocalizeSpanishDialect",
     "LocalizeTargetLanguage",
     "LocalizeVoiceRequest",
     "LocalizeVoiceRequestParams",

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/src/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": "2.0.0b7",
+            "X-Fern-SDK-Version": "2.0.0b8",
         }
         headers["X-API-Key"] = self.api_key
         headers["Cartesia-Version"] = "2024-11-13"

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/src/cartesia/tts/_async_websocket.py

@@ -69,6 +69,7 @@ class _AsyncTTSContext:
         stream: bool = True,
         add_timestamps: bool = False,
         add_phoneme_timestamps: bool = False,
+        use_original_timestamps: bool = False,
         continue_: bool = False,
         flush: bool = False,
     ) -> None:
@@ -106,6 +107,8 @@ class _AsyncTTSContext:
             request_body["add_timestamps"] = add_timestamps
         if add_phoneme_timestamps:
             request_body["add_phoneme_timestamps"] = add_phoneme_timestamps
+        if use_original_timestamps:
+            request_body["use_original_timestamps"] = use_original_timestamps
         if continue_:
             request_body["continue"] = continue_
         if flush:
@@ -367,6 +370,7 @@ class AsyncTtsWebsocket(TtsWebsocket):
         stream: bool = True,
         add_timestamps: bool = False,
         add_phoneme_timestamps: bool = False,
+        use_original_timestamps: bool = False,
     ):
         """See :meth:`_WebSocket.send` for details."""
         if context_id is None:
@@ -385,6 +389,7 @@ class AsyncTtsWebsocket(TtsWebsocket):
             continue_=False,
             add_timestamps=add_timestamps,
             add_phoneme_timestamps=add_phoneme_timestamps,
+            use_original_timestamps=use_original_timestamps,
         )
 
         generator = ctx.receive()

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/src/cartesia/tts/_websocket.py

@@ -67,6 +67,8 @@ class _TTSContext:
         language: Optional[str] = None,
         stream: bool = True,
         add_timestamps: bool = False,
+        add_phoneme_timestamps: bool = False,
+        use_original_timestamps: bool = False,
     ) -> Generator[bytes, None, None]:
         """Send audio generation requests to the WebSocket and yield responses.
 
@@ -102,6 +104,10 @@ class _TTSContext:
             request_body["stream"] = stream
         if add_timestamps:
             request_body["add_timestamps"] = add_timestamps
+        if add_phoneme_timestamps:
+            request_body["add_phoneme_timestamps"] = add_phoneme_timestamps
+        if use_original_timestamps:
+            request_body["use_original_timestamps"] = use_original_timestamps
 
         if (
             "context_id" in request_body
@@ -354,6 +360,7 @@ class TtsWebsocket:
         stream: bool = True,
         add_timestamps: bool = False,
         add_phoneme_timestamps: bool = False,
+        use_original_timestamps: bool = False,
     ):
         """Send a request to the WebSocket to generate audio.
 
@@ -384,6 +391,7 @@ class TtsWebsocket:
             "stream": stream,
             "add_timestamps": add_timestamps,
             "add_phoneme_timestamps": add_phoneme_timestamps,
+            "use_original_timestamps": use_original_timestamps,
         }
         generator = self._websocket_generator(request_body)
 

{cartesia-2.0.0b7 → cartesia-2.0.0b8}/src/cartesia/voices/__init__.py

@@ -12,6 +12,9 @@ from .types import (
     IdSpecifier,
     LocalizeDialect,
     LocalizeEnglishDialect,
+    LocalizeFrenchDialect,
+    LocalizePortugueseDialect,
+    LocalizeSpanishDialect,
     LocalizeTargetLanguage,
     LocalizeVoiceRequest,
     MixVoiceSpecifier,
@@ -56,6 +59,9 @@ __all__ = [
     "LocalizeDialect",
     "LocalizeDialectParams",
     "LocalizeEnglishDialect",
+    "LocalizeFrenchDialect",
+    "LocalizePortugueseDialect",
+    "LocalizeSpanishDialect",
     "LocalizeTargetLanguage",
     "LocalizeVoiceRequest",
     "LocalizeVoiceRequestParams",