sarvamai 0.1.23a4__py3-none-any.whl → 0.1.23a6__py3-none-any.whl

sarvamai/__init__.py

@@ -188,6 +188,7 @@ from .requests import (
 from .speech_to_text_streaming import (
     SpeechToTextStreamingFlushSignal,
     SpeechToTextStreamingHighVadSensitivity,
+    SpeechToTextStreamingInputAudioCodec,
     SpeechToTextStreamingLanguageCode,
     SpeechToTextStreamingMode,
     SpeechToTextStreamingModel,
@@ -196,6 +197,7 @@ from .speech_to_text_streaming import (
 from .speech_to_text_translate_streaming import (
     SpeechToTextTranslateStreamingFlushSignal,
     SpeechToTextTranslateStreamingHighVadSensitivity,
+    SpeechToTextTranslateStreamingInputAudioCodec,
     SpeechToTextTranslateStreamingVadSignals,
 )
 from .text_to_speech_streaming import TextToSpeechStreamingModel, TextToSpeechStreamingSendCompletionEvent
@@ -320,6 +322,7 @@ __all__ = [
     "SpeechToTextResponseParams",
     "SpeechToTextStreamingFlushSignal",
     "SpeechToTextStreamingHighVadSensitivity",
+    "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingMode",
     "SpeechToTextStreamingModel",
@@ -338,6 +341,7 @@ __all__ = [
     "SpeechToTextTranslateResponseParams",
     "SpeechToTextTranslateStreamingFlushSignal",
     "SpeechToTextTranslateStreamingHighVadSensitivity",
+    "SpeechToTextTranslateStreamingInputAudioCodec",
     "SpeechToTextTranslateStreamingResponse",
     "SpeechToTextTranslateStreamingResponseParams",
     "SpeechToTextTranslateStreamingVadSignals",

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/0.1.23a6",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.23a4",
+            "X-Fern-SDK-Version": "0.1.23a6",
             **(self.get_custom_headers() or {}),
         }
         headers["api-subscription-key"] = self.api_subscription_key

sarvamai/requests/speech_to_text_job_parameters.py

@@ -9,20 +9,52 @@ from ..types.speech_to_text_translate_language import SpeechToTextTranslateLangu
 class SpeechToTextJobParametersParams(typing_extensions.TypedDict):
     language_code: typing_extensions.NotRequired[SpeechToTextTranslateLanguage]
     """
-    Language code
+    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)
-    - **saarika:v3**: Advanced transcription model
-    - **saaras:v3**: Advanced model with multiple 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_extensions.NotRequired[Mode]
     """
-    Mode of operation. Only applicable for saaras:v3 model.
+    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]

sarvamai/requests/speech_to_text_response.py

@@ -1,28 +1,20 @@
 # This file was auto-generated by Fern from our API Definition.
 
 import typing_extensions
-from .diarized_transcript import DiarizedTranscriptParams
-from .timestamps_model import TimestampsModelParams
 
 
 class SpeechToTextResponseParams(typing_extensions.TypedDict):
-    request_id: typing_extensions.NotRequired[str]
-    transcript: str
+    request_id: str
     """
-    The transcribed text from the provided audio file.
+    Unique identifier for the request
     """
 
-    timestamps: typing_extensions.NotRequired[TimestampsModelParams]
-    """
-    Contains timestamps for the transcribed text. This field is included only if with_timestamps is set to true
-    """
-
-    diarized_transcript: typing_extensions.NotRequired[DiarizedTranscriptParams]
+    transcript: str
     """
-    Diarized transcript of the provided speech
+    The transcribed text from the provided audio file.
     """
 
-    language_code: typing_extensions.NotRequired[str]
+    language_code: str
     """
-    This will return the BCP-47 code of language spoken in the input. If multiple languages are detected, this will return language code of most predominant spoken language. If no language is detected, this will be null
+    The BCP-47 code of language spoken in the input (e.g., hi-IN, en-IN). If multiple languages are detected, returns the most predominant spoken language.
     """

sarvamai/requests/speech_to_text_transcription_data.py

@@ -32,4 +32,18 @@ class SpeechToTextTranscriptionDataParams(typing_extensions.TypedDict):
     BCP-47 code of detected language
     """
 
+    language_probability: typing_extensions.NotRequired[float]
+    """
+    Float value (0.0 to 1.0) indicating the probability of the detected language being correct. Higher values indicate higher confidence.
+    
+    **When it returns a value:**
+    - When `language_code` is not provided in the request
+    - When `language_code` is set to `unknown`
+    
+    **When it returns null:**
+    - When a specific `language_code` is provided (language detection is skipped)
+    
+    The parameter is always present in the response.
+    """
+
     metrics: TranscriptionMetricsParams

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/requests/speech_to_text_translate_response.py

@@ -1,23 +1,20 @@
 # This file was auto-generated by Fern from our API Definition.
 
 import typing_extensions
-from ..types.speech_to_text_translate_language import SpeechToTextTranslateLanguage
-from .diarized_transcript import DiarizedTranscriptParams
 
 
 class SpeechToTextTranslateResponseParams(typing_extensions.TypedDict):
-    request_id: typing_extensions.NotRequired[str]
-    transcript: str
+    request_id: str
     """
-    Transcript of the provided speech
+    Unique identifier for the request
     """
 
-    language_code: typing_extensions.NotRequired[SpeechToTextTranslateLanguage]
+    transcript: str
     """
-    This will return the BCP-47 code of language spoken in the input. If multiple languages are detected, this will return language code of most predominant spoken language. If no language is detected, this will be null
+    English translation of the provided speech
     """
 
-    diarized_transcript: typing_extensions.NotRequired[DiarizedTranscriptParams]
+    language_code: str
     """
-    Diarized transcript of the provided speech
+    The BCP-47 code of the detected source language spoken in the input (e.g., hi-IN, kn-IN).
     """

sarvamai/requests/speech_to_text_translate_transcription_data.py

@@ -20,4 +20,17 @@ class SpeechToTextTranslateTranscriptionDataParams(typing_extensions.TypedDict):
     BCP-47 code of detected source language (null when language detection is in progress)
     """
 
+    language_probability: typing_extensions.NotRequired[float]
+    """
+    Float value (0.0 to 1.0) indicating the probability of the detected source language being correct. Higher values indicate higher confidence.
+    
+    **When it returns a value:**
+    - Always returns a value as source language is auto-detected for translation
+    
+    **When it returns null:**
+    - When language detection confidence is unavailable
+    
+    The parameter is always present in the response.
+    """
+
     metrics: TranscriptionMetricsParams

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/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/speech_to_text_streaming/__init__.py

@@ -5,6 +5,7 @@
 from .types import (
     SpeechToTextStreamingFlushSignal,
     SpeechToTextStreamingHighVadSensitivity,
+    SpeechToTextStreamingInputAudioCodec,
     SpeechToTextStreamingLanguageCode,
     SpeechToTextStreamingMode,
     SpeechToTextStreamingModel,
@@ -14,6 +15,7 @@ from .types import (
 __all__ = [
     "SpeechToTextStreamingFlushSignal",
     "SpeechToTextStreamingHighVadSensitivity",
+    "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
     "SpeechToTextStreamingMode",
     "SpeechToTextStreamingModel",