sarvamai 0.1.22a4__py3-none-any.whl → 0.1.22a8__py3-none-any.whl

sarvamai/speech_to_text_streaming/client.py

@@ -15,6 +15,8 @@ from .types.speech_to_text_streaming_flush_signal import SpeechToTextStreamingFl
 from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 from .types.speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
+from .types.speech_to_text_streaming_mode import SpeechToTextStreamingMode
+from .types.speech_to_text_streaming_model import SpeechToTextStreamingModel
 from .types.speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
 try:
@@ -43,12 +45,13 @@ class SpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
+        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        mode: typing.Optional[SpeechToTextStreamingMode] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -62,13 +65,47 @@ class SpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition
+            Specifies the language of the input audio in BCP-47 format.
 
-        model : typing.Optional[typing.Literal["saarika:v2.5"]]
-            Speech to text model to use
+            **Available Options:**
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `gu-IN`: Gujarati
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
 
-        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+        model : typing.Optional[SpeechToTextStreamingModel]
+            Specifies the model to use for speech-to-text conversion.
+
+            - **saarika:v2.5** (default): Transcribes audio in the spoken language.
+
+            - **saaras:v3**: State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
+
+        mode : typing.Optional[SpeechToTextStreamingMode]
+            Mode of operation. **Only applicable when using saaras:v3 model.**
+
+            Example audio: 'मेरा फोन नंबर है 9840950950'
+
+            - **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.
+              - Output: `मेरा फोन नंबर है 9840950950`
+
+            - **translate**: Translates speech from any supported Indic language to English.
+              - Output: `My phone number is 9840950950`
+
+            - **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.
+              - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`
+
+            - **translit**: Romanization - Transliterates speech to Latin/Roman script only.
+              - Output: `mera phone number hai 9840950950`
+
+            - **codemix**: Code-mixed text with English words in English and Indic words in native script.
+              - Output: `मेरा phone number है 9840950950`
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -82,6 +119,10 @@ class SpeechToTextStreamingClient:
         flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription
 
+        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -98,8 +139,8 @@ class SpeechToTextStreamingClient:
             query_params = query_params.add("language-code", language_code)
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
+        if mode is not None:
+            query_params = query_params.add("mode", mode)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -108,6 +149,8 @@ class SpeechToTextStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -152,12 +195,13 @@ class AsyncSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
+        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        mode: typing.Optional[SpeechToTextStreamingMode] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -171,13 +215,47 @@ class AsyncSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition
+            Specifies the language of the input audio in BCP-47 format.
 
-        model : typing.Optional[typing.Literal["saarika:v2.5"]]
-            Speech to text model to use
+            **Available Options:**
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `gu-IN`: Gujarati
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
 
-        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+        model : typing.Optional[SpeechToTextStreamingModel]
+            Specifies the model to use for speech-to-text conversion.
+
+            - **saarika:v2.5** (default): Transcribes audio in the spoken language.
+
+            - **saaras:v3**: State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
+
+        mode : typing.Optional[SpeechToTextStreamingMode]
+            Mode of operation. **Only applicable when using saaras:v3 model.**
+
+            Example audio: 'मेरा फोन नंबर है 9840950950'
+
+            - **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.
+              - Output: `मेरा फोन नंबर है 9840950950`
+
+            - **translate**: Translates speech from any supported Indic language to English.
+              - Output: `My phone number is 9840950950`
+
+            - **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.
+              - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`
+
+            - **translit**: Romanization - Transliterates speech to Latin/Roman script only.
+              - Output: `mera phone number hai 9840950950`
+
+            - **codemix**: Code-mixed text with English words in English and Indic words in native script.
+              - Output: `मेरा phone number है 9840950950`
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -191,6 +269,10 @@ class AsyncSpeechToTextStreamingClient:
         flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription
 
+        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -207,8 +289,8 @@ class AsyncSpeechToTextStreamingClient:
             query_params = query_params.add("language-code", language_code)
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
+        if mode is not None:
+            query_params = query_params.add("mode", mode)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -217,6 +299,8 @@ class AsyncSpeechToTextStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:

sarvamai/speech_to_text_streaming/raw_client.py

@@ -14,6 +14,8 @@ from .types.speech_to_text_streaming_flush_signal import SpeechToTextStreamingFl
 from .types.speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .types.speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 from .types.speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
+from .types.speech_to_text_streaming_mode import SpeechToTextStreamingMode
+from .types.speech_to_text_streaming_model import SpeechToTextStreamingModel
 from .types.speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
 try:
@@ -31,12 +33,13 @@ class RawSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
+        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        mode: typing.Optional[SpeechToTextStreamingMode] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextStreamingSocketClient]:
@@ -50,13 +53,47 @@ class RawSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition
+            Specifies the language of the input audio in BCP-47 format.
 
-        model : typing.Optional[typing.Literal["saarika:v2.5"]]
-            Speech to text model to use
+            **Available Options:**
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `gu-IN`: Gujarati
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
 
-        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+        model : typing.Optional[SpeechToTextStreamingModel]
+            Specifies the model to use for speech-to-text conversion.
+
+            - **saarika:v2.5** (default): Transcribes audio in the spoken language.
+
+            - **saaras:v3**: State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
+
+        mode : typing.Optional[SpeechToTextStreamingMode]
+            Mode of operation. **Only applicable when using saaras:v3 model.**
+
+            Example audio: 'मेरा फोन नंबर है 9840950950'
+
+            - **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.
+              - Output: `मेरा फोन नंबर है 9840950950`
+
+            - **translate**: Translates speech from any supported Indic language to English.
+              - Output: `My phone number is 9840950950`
+
+            - **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.
+              - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`
+
+            - **translit**: Romanization - Transliterates speech to Latin/Roman script only.
+              - Output: `mera phone number hai 9840950950`
+
+            - **codemix**: Code-mixed text with English words in English and Indic words in native script.
+              - Output: `मेरा phone number है 9840950950`
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -70,6 +107,10 @@ class RawSpeechToTextStreamingClient:
         flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription
 
+        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -86,8 +127,8 @@ class RawSpeechToTextStreamingClient:
             query_params = query_params.add("language-code", language_code)
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
+        if mode is not None:
+            query_params = query_params.add("mode", mode)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -96,6 +137,8 @@ class RawSpeechToTextStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -129,12 +172,13 @@ class AsyncRawSpeechToTextStreamingClient:
         self,
         *,
         language_code: SpeechToTextStreamingLanguageCode,
-        model: typing.Optional[typing.Literal["saarika:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
+        model: typing.Optional[SpeechToTextStreamingModel] = None,
+        mode: typing.Optional[SpeechToTextStreamingMode] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextStreamingSocketClient]:
@@ -148,13 +192,47 @@ class AsyncRawSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition
+            Specifies the language of the input audio in BCP-47 format.
 
-        model : typing.Optional[typing.Literal["saarika:v2.5"]]
-            Speech to text model to use
+            **Available Options:**
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `gu-IN`: Gujarati
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
 
-        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+        model : typing.Optional[SpeechToTextStreamingModel]
+            Specifies the model to use for speech-to-text conversion.
+
+            - **saarika:v2.5** (default): Transcribes audio in the spoken language.
+
+            - **saaras:v3**: State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
+
+        mode : typing.Optional[SpeechToTextStreamingMode]
+            Mode of operation. **Only applicable when using saaras:v3 model.**
+
+            Example audio: 'मेरा फोन नंबर है 9840950950'
+
+            - **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.
+              - Output: `मेरा फोन नंबर है 9840950950`
+
+            - **translate**: Translates speech from any supported Indic language to English.
+              - Output: `My phone number is 9840950950`
+
+            - **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.
+              - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`
+
+            - **translit**: Romanization - Transliterates speech to Latin/Roman script only.
+              - Output: `mera phone number hai 9840950950`
+
+            - **codemix**: Code-mixed text with English words in English and Indic words in native script.
+              - Output: `मेरा phone number है 9840950950`
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -168,6 +246,10 @@ class AsyncRawSpeechToTextStreamingClient:
         flush_signal : typing.Optional[SpeechToTextStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription
 
+        input_audio_codec : typing.Optional[SpeechToTextStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -184,8 +266,8 @@ class AsyncRawSpeechToTextStreamingClient:
             query_params = query_params.add("language-code", language_code)
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
+        if mode is not None:
+            query_params = query_params.add("mode", mode)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -194,6 +276,8 @@ class AsyncRawSpeechToTextStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._client_wrapper.get_headers()
         if api_subscription_key is not None:

sarvamai/speech_to_text_streaming/types/__init__.py

@@ -6,6 +6,8 @@ from .speech_to_text_streaming_flush_signal import SpeechToTextStreamingFlushSig
 from .speech_to_text_streaming_high_vad_sensitivity import SpeechToTextStreamingHighVadSensitivity
 from .speech_to_text_streaming_input_audio_codec import SpeechToTextStreamingInputAudioCodec
 from .speech_to_text_streaming_language_code import SpeechToTextStreamingLanguageCode
+from .speech_to_text_streaming_mode import SpeechToTextStreamingMode
+from .speech_to_text_streaming_model import SpeechToTextStreamingModel
 from .speech_to_text_streaming_vad_signals import SpeechToTextStreamingVadSignals
 
 __all__ = [
@@ -13,5 +15,7 @@ __all__ = [
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
+    "SpeechToTextStreamingMode",
+    "SpeechToTextStreamingModel",
     "SpeechToTextStreamingVadSignals",
 ]

sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_input_audio_codec.py

@@ -3,31 +3,5 @@
 import typing
 
 SpeechToTextStreamingInputAudioCodec = typing.Union[
-    typing.Literal[
-        "wav",
-        "x-wav",
-        "wave",
-        "mp3",
-        "mpeg",
-        "mpeg3",
-        "x-mp3",
-        "x-mpeg-3",
-        "aac",
-        "x-aac",
-        "aiff",
-        "x-aiff",
-        "ogg",
-        "opus",
-        "flac",
-        "x-flac",
-        "mp4",
-        "x-m4a",
-        "amr",
-        "x-ms-wma",
-        "webm",
-        "pcm_s16le",
-        "pcm_l16",
-        "pcm_raw",
-    ],
-    typing.Any,
+    typing.Literal["wav", "pcm_s16le", "pcm_l16", "pcm_raw"], typing.Any
 ]

sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_mode.py

@@ -0,0 +1,7 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextStreamingMode = typing.Union[
+    typing.Literal["transcribe", "translate", "verbatim", "translit", "codemix"], typing.Any
+]

sarvamai/speech_to_text_streaming/types/speech_to_text_streaming_model.py

@@ -0,0 +1,5 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing
+
+SpeechToTextStreamingModel = typing.Union[typing.Literal["saarika:v2.5", "saaras:v3"], typing.Any]

sarvamai/speech_to_text_translate_streaming/client.py

@@ -44,11 +44,11 @@ class SpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.Iterator[SpeechToTextTranslateStreamingSocketClient]:
@@ -62,10 +62,10 @@ class SpeechToTextTranslateStreamingClient:
         Parameters
         ----------
         model : typing.Optional[typing.Literal["saaras:v2.5"]]
-            Speech to text model to use (defaults to "saaras:v2.5" if not specified)
+            Model to be used for speech to text translation.
 
-        input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -79,6 +79,10 @@ class SpeechToTextTranslateStreamingClient:
         flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription and translation
 
+        input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -93,8 +97,6 @@ class SpeechToTextTranslateStreamingClient:
         query_params = httpx.QueryParams()
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -103,6 +105,8 @@ class SpeechToTextTranslateStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None:
@@ -147,11 +151,11 @@ class AsyncSpeechToTextTranslateStreamingClient:
         self,
         *,
         model: typing.Optional[typing.Literal["saaras:v2.5"]] = None,
-        input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         sample_rate: typing.Optional[str] = None,
         high_vad_sensitivity: typing.Optional[SpeechToTextTranslateStreamingHighVadSensitivity] = None,
         vad_signals: typing.Optional[SpeechToTextTranslateStreamingVadSignals] = None,
         flush_signal: typing.Optional[SpeechToTextTranslateStreamingFlushSignal] = None,
+        input_audio_codec: typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec] = None,
         api_subscription_key: typing.Optional[str] = None,
         request_options: typing.Optional[RequestOptions] = None,
     ) -> typing.AsyncIterator[AsyncSpeechToTextTranslateStreamingSocketClient]:
@@ -165,10 +169,10 @@ class AsyncSpeechToTextTranslateStreamingClient:
         Parameters
         ----------
         model : typing.Optional[typing.Literal["saaras:v2.5"]]
-            Speech to text model to use (defaults to "saaras:v2.5" if not specified)
+            Model to be used for speech to text translation.
 
-        input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
-            Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files supports sample rate 16000 and 8000.
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         sample_rate : typing.Optional[str]
             Audio sample rate for the WebSocket connection. When specified as a connection parameter, only 16kHz and 8kHz are supported. 8kHz is only available via this connection parameter. If not specified, defaults to 16kHz.
@@ -182,6 +186,10 @@ class AsyncSpeechToTextTranslateStreamingClient:
         flush_signal : typing.Optional[SpeechToTextTranslateStreamingFlushSignal]
             Signal to flush the audio buffer and finalize transcription and translation
 
+        input_audio_codec : typing.Optional[SpeechToTextTranslateStreamingInputAudioCodec]
+            Audio codec/format of the input stream. Use this when sending raw PCM audio.
+            Supported values: wav, pcm_s16le, pcm_l16, pcm_raw.
+
         api_subscription_key : typing.Optional[str]
             API subscription key for authentication
 
@@ -196,8 +204,6 @@ class AsyncSpeechToTextTranslateStreamingClient:
         query_params = httpx.QueryParams()
         if model is not None:
             query_params = query_params.add("model", model)
-        if input_audio_codec is not None:
-            query_params = query_params.add("input_audio_codec", input_audio_codec)
         if sample_rate is not None:
             query_params = query_params.add("sample_rate", sample_rate)
         if high_vad_sensitivity is not None:
@@ -206,6 +212,8 @@ class AsyncSpeechToTextTranslateStreamingClient:
             query_params = query_params.add("vad_signals", vad_signals)
         if flush_signal is not None:
             query_params = query_params.add("flush_signal", flush_signal)
+        if input_audio_codec is not None:
+            query_params = query_params.add("input_audio_codec", input_audio_codec)
         ws_url = ws_url + f"?{query_params}"
         headers = self._raw_client._client_wrapper.get_headers()
         if api_subscription_key is not None: