sarvamai 0.1.23a4__tar.gz → 0.1.23a5__tar.gz

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.23a4
+Version: 0.1.23a5
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/pyproject.toml

@@ -3,7 +3,7 @@ name = "sarvamai"
 
 [tool.poetry]
 name = "sarvamai"
-version = "0.1.23a4"
+version = "0.1.23a5"
 description = ""
 readme = "README.md"
 authors = []

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/core/client_wrapper.py

@@ -23,10 +23,10 @@ class BaseClientWrapper:
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.23a4",
+            "User-Agent": "sarvamai/0.1.23a5",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.23a4",
+            "X-Fern-SDK-Version": "0.1.23a5",
             **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key

sarvamai-0.1.23a5/src/sarvamai/requests/speech_to_text_job_parameters.py

@@ -0,0 +1,73 @@
+# This file was auto-generated by Fern from our API Definition.
+
+import typing_extensions
+from ..types.mode import Mode
+from ..types.speech_to_text_model import SpeechToTextModel
+from ..types.speech_to_text_translate_language import SpeechToTextTranslateLanguage
+
+
+class SpeechToTextJobParametersParams(typing_extensions.TypedDict):
+    language_code: typing_extensions.NotRequired[SpeechToTextTranslateLanguage]
+    """
+    Specifies the language of the input audio in BCP-47 format.
+    
+    **Available Options:**
+    - `unknown` (default): Use when the language is not known; the API will auto-detect.
+    - `hi-IN`: Hindi
+    - `bn-IN`: Bengali
+    - `kn-IN`: Kannada
+    - `ml-IN`: Malayalam
+    - `mr-IN`: Marathi
+    - `od-IN`: Odia
+    - `pa-IN`: Punjabi
+    - `ta-IN`: Tamil
+    - `te-IN`: Telugu
+    - `en-IN`: English
+    - `gu-IN`: Gujarati
+    """
+
+    model: typing_extensions.NotRequired[SpeechToTextModel]
+    """
+    Model to be used for speech to text.
+    
+    - **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_extensions.NotRequired[Mode]
+    """
+    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`
+    """
+
+    with_timestamps: typing_extensions.NotRequired[bool]
+    """
+    Whether to include timestamps in the response
+    """
+
+    with_diarization: typing_extensions.NotRequired[bool]
+    """
+    Enables speaker diarization, which identifies and separates different speakers in the audio. In beta mode
+    """
+
+    num_speakers: typing_extensions.NotRequired[int]
+    """
+    Number of speakers to be detected in the audio. This is used when with_diarization is true.
+    """

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/requests/speech_to_text_translate_job_parameters.py

@@ -12,7 +12,10 @@ class SpeechToTextTranslateJobParametersParams(typing_extensions.TypedDict):
 
     model: typing_extensions.NotRequired[SpeechToTextTranslateModel]
     """
-    Model to be used for converting speech to text in target language
+    Model to be used for speech to text translation.
+    
+    - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+      - Example: Hindi audio → English text output
     """
 
     with_diarization: typing_extensions.NotRequired[bool]

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/speech_to_text/client.py

@@ -65,23 +65,49 @@ class SpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            - **saarika:v2.5** (default): Standard transcription model
-            - **saarika:v3**: Advanced transcription model
-            - **saaras:v3**: Advanced model with multiple output modes
+
+            - **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[Mode]
             Mode of operation. **Only applicable when using saaras:v3 model.**
-            - **transcribe** (default): Standard transcription
-            - **translate**: Translation to English
-            - **indic-en**: Indic to English translation
-            - **verbatim**: Exact transcription
-            - **translit**: Transliteration to Latin script
-            - **codemix**: Code-mixed output
+
+            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`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio.
-             For the `saarika:v2.5` model, it is optional.
-            `unknown`: Use this when the language is not known; the API will detect it automatically.
+            Specifies the language of the input audio in BCP-47 format.
+
+            **Note:** This parameter is optional for `saarika:v2.5` model.
+
+            **Available Options:**
+            - `unknown`: Use when the language is not known; the API will auto-detect.
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
+            - `gu-IN`: Gujarati
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate.
@@ -146,7 +172,10 @@ class SpeechToTextClient:
             Conversation context can be passed as a prompt to boost model accuracy. However, the current system is at an experimentation stage and doesn't match the prompt performance of large language models.
 
         model : typing.Optional[SpeechToTextTranslateModel]
-            Model to be used for converting speech to text in target language
+            Model to be used for speech to text translation.
+
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             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 are supported only at 16kHz sample rate.
@@ -221,23 +250,49 @@ class AsyncSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            - **saarika:v2.5** (default): Standard transcription model
-            - **saarika:v3**: Advanced transcription model
-            - **saaras:v3**: Advanced model with multiple output modes
+
+            - **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[Mode]
             Mode of operation. **Only applicable when using saaras:v3 model.**
-            - **transcribe** (default): Standard transcription
-            - **translate**: Translation to English
-            - **indic-en**: Indic to English translation
-            - **verbatim**: Exact transcription
-            - **translit**: Transliteration to Latin script
-            - **codemix**: Code-mixed output
+
+            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`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio.
-             For the `saarika:v2.5` model, it is optional.
-            `unknown`: Use this when the language is not known; the API will detect it automatically.
+            Specifies the language of the input audio in BCP-47 format.
+
+            **Note:** This parameter is optional for `saarika:v2.5` model.
+
+            **Available Options:**
+            - `unknown`: Use when the language is not known; the API will auto-detect.
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
+            - `gu-IN`: Gujarati
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate.
@@ -310,7 +365,10 @@ class AsyncSpeechToTextClient:
             Conversation context can be passed as a prompt to boost model accuracy. However, the current system is at an experimentation stage and doesn't match the prompt performance of large language models.
 
         model : typing.Optional[SpeechToTextTranslateModel]
-            Model to be used for converting speech to text in target language
+            Model to be used for speech to text translation.
+
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             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 are supported only at 16kHz sample rate.

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/speech_to_text/raw_client.py

@@ -63,23 +63,49 @@ class RawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            - **saarika:v2.5** (default): Standard transcription model
-            - **saarika:v3**: Advanced transcription model
-            - **saaras:v3**: Advanced model with multiple output modes
+
+            - **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[Mode]
             Mode of operation. **Only applicable when using saaras:v3 model.**
-            - **transcribe** (default): Standard transcription
-            - **translate**: Translation to English
-            - **indic-en**: Indic to English translation
-            - **verbatim**: Exact transcription
-            - **translit**: Transliteration to Latin script
-            - **codemix**: Code-mixed output
+
+            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`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio.
-             For the `saarika:v2.5` model, it is optional.
-            `unknown`: Use this when the language is not known; the API will detect it automatically.
+            Specifies the language of the input audio in BCP-47 format.
+
+            **Note:** This parameter is optional for `saarika:v2.5` model.
+
+            **Available Options:**
+            - `unknown`: Use when the language is not known; the API will auto-detect.
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
+            - `gu-IN`: Gujarati
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate.
@@ -223,7 +249,10 @@ class RawSpeechToTextClient:
             Conversation context can be passed as a prompt to boost model accuracy. However, the current system is at an experimentation stage and doesn't match the prompt performance of large language models.
 
         model : typing.Optional[SpeechToTextTranslateModel]
-            Model to be used for converting speech to text in target language
+            Model to be used for speech to text translation.
+
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             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 are supported only at 16kHz sample rate.
@@ -370,23 +399,49 @@ class AsyncRawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            - **saarika:v2.5** (default): Standard transcription model
-            - **saarika:v3**: Advanced transcription model
-            - **saaras:v3**: Advanced model with multiple output modes
+
+            - **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[Mode]
             Mode of operation. **Only applicable when using saaras:v3 model.**
-            - **transcribe** (default): Standard transcription
-            - **translate**: Translation to English
-            - **indic-en**: Indic to English translation
-            - **verbatim**: Exact transcription
-            - **translit**: Transliteration to Latin script
-            - **codemix**: Code-mixed output
+
+            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`
 
         language_code : typing.Optional[SpeechToTextLanguage]
-            Specifies the language of the input audio.
-             For the `saarika:v2.5` model, it is optional.
-            `unknown`: Use this when the language is not known; the API will detect it automatically.
+            Specifies the language of the input audio in BCP-47 format.
+
+            **Note:** This parameter is optional for `saarika:v2.5` model.
+
+            **Available Options:**
+            - `unknown`: Use when the language is not known; the API will auto-detect.
+            - `hi-IN`: Hindi
+            - `bn-IN`: Bengali
+            - `kn-IN`: Kannada
+            - `ml-IN`: Malayalam
+            - `mr-IN`: Marathi
+            - `od-IN`: Odia
+            - `pa-IN`: Punjabi
+            - `ta-IN`: Tamil
+            - `te-IN`: Telugu
+            - `en-IN`: English
+            - `gu-IN`: Gujarati
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate.
@@ -530,7 +585,10 @@ class AsyncRawSpeechToTextClient:
             Conversation context can be passed as a prompt to boost model accuracy. However, the current system is at an experimentation stage and doesn't match the prompt performance of large language models.
 
         model : typing.Optional[SpeechToTextTranslateModel]
-            Model to be used for converting speech to text in target language
+            Model to be used for speech to text translation.
+
+            - **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.
+              - Example: Hindi audio → English text output
 
         input_audio_codec : typing.Optional[InputAudioCodec]
             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 are supported only at 16kHz sample rate.

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/speech_to_text_streaming/client.py

@@ -63,19 +63,47 @@ class SpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition (BCP-47 format)
+            Specifies the language of the input audio in BCP-47 format.
+
+            **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
 
         model : typing.Optional[SpeechToTextStreamingModel]
-            Speech to text model to use
+            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 for saaras:v3 model. Only applicable when model is 'saaras:v3'.
-            - transcribe: Standard Whisper transcription
-            - translate: Standard Whisper translation to English
-            - indic-en: Translate Indic languages to English
-            - verbatim: Exact transcription in original script
-            - translit: Transliteration to Latin script
-            - codemix: Code-mixed output (native + English)
+            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.
@@ -178,19 +206,47 @@ class AsyncSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition (BCP-47 format)
+            Specifies the language of the input audio in BCP-47 format.
+
+            **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
 
         model : typing.Optional[SpeechToTextStreamingModel]
-            Speech to text model to use
+            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 for saaras:v3 model. Only applicable when model is 'saaras:v3'.
-            - transcribe: Standard Whisper transcription
-            - translate: Standard Whisper translation to English
-            - indic-en: Translate Indic languages to English
-            - verbatim: Exact transcription in original script
-            - translit: Transliteration to Latin script
-            - codemix: Code-mixed output (native + English)
+            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.

{sarvamai-0.1.23a4 → sarvamai-0.1.23a5}/src/sarvamai/speech_to_text_streaming/raw_client.py

@@ -51,19 +51,47 @@ class RawSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition (BCP-47 format)
+            Specifies the language of the input audio in BCP-47 format.
+
+            **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
 
         model : typing.Optional[SpeechToTextStreamingModel]
-            Speech to text model to use
+            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 for saaras:v3 model. Only applicable when model is 'saaras:v3'.
-            - transcribe: Standard Whisper transcription
-            - translate: Standard Whisper translation to English
-            - indic-en: Translate Indic languages to English
-            - verbatim: Exact transcription in original script
-            - translit: Transliteration to Latin script
-            - codemix: Code-mixed output (native + English)
+            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.
@@ -155,19 +183,47 @@ class AsyncRawSpeechToTextStreamingClient:
         Parameters
         ----------
         language_code : SpeechToTextStreamingLanguageCode
-            Language code for speech recognition (BCP-47 format)
+            Specifies the language of the input audio in BCP-47 format.
+
+            **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
 
         model : typing.Optional[SpeechToTextStreamingModel]
-            Speech to text model to use
+            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 for saaras:v3 model. Only applicable when model is 'saaras:v3'.
-            - transcribe: Standard Whisper transcription
-            - translate: Standard Whisper translation to English
-            - indic-en: Translate Indic languages to English
-            - verbatim: Exact transcription in original script
-            - translit: Transliteration to Latin script
-            - codemix: Code-mixed output (native + English)
+            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.