sarvamai 0.1.22a4__py3-none-any.whl → 0.1.22a7__py3-none-any.whl

sarvamai/speech_to_text/client.py

@@ -6,6 +6,7 @@ from .. import core
 from ..core.client_wrapper import AsyncClientWrapper, SyncClientWrapper
 from ..core.request_options import RequestOptions
 from ..types.input_audio_codec import InputAudioCodec
+from ..types.mode import Mode
 from ..types.speech_to_text_language import SpeechToTextLanguage
 from ..types.speech_to_text_model import SpeechToTextModel
 from ..types.speech_to_text_response import SpeechToTextResponse
@@ -37,6 +38,7 @@ class SpeechToTextClient:
         *,
         file: core.File,
         model: typing.Optional[SpeechToTextModel] = OMIT,
+        mode: typing.Optional[Mode] = OMIT,
         language_code: typing.Optional[SpeechToTextLanguage] = OMIT,
         input_audio_codec: typing.Optional[InputAudioCodec] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
@@ -63,12 +65,49 @@ class SpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2.5`
+
+            - **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.**
+
+            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.
@@ -93,6 +132,7 @@ class SpeechToTextClient:
         _response = self._raw_client.transcribe(
             file=file,
             model=model,
+            mode=mode,
             language_code=language_code,
             input_audio_codec=input_audio_codec,
             request_options=request_options,
@@ -132,7 +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.
@@ -180,6 +223,7 @@ class AsyncSpeechToTextClient:
         *,
         file: core.File,
         model: typing.Optional[SpeechToTextModel] = OMIT,
+        mode: typing.Optional[Mode] = OMIT,
         language_code: typing.Optional[SpeechToTextLanguage] = OMIT,
         input_audio_codec: typing.Optional[InputAudioCodec] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
@@ -206,12 +250,49 @@ class AsyncSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2.5`
+
+            - **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.**
+
+            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.
@@ -244,6 +325,7 @@ class AsyncSpeechToTextClient:
         _response = await self._raw_client.transcribe(
             file=file,
             model=model,
+            mode=mode,
             language_code=language_code,
             input_audio_codec=input_audio_codec,
             request_options=request_options,
@@ -283,7 +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

@@ -16,6 +16,7 @@ from ..errors.service_unavailable_error import ServiceUnavailableError
 from ..errors.too_many_requests_error import TooManyRequestsError
 from ..errors.unprocessable_entity_error import UnprocessableEntityError
 from ..types.input_audio_codec import InputAudioCodec
+from ..types.mode import Mode
 from ..types.speech_to_text_language import SpeechToTextLanguage
 from ..types.speech_to_text_model import SpeechToTextModel
 from ..types.speech_to_text_response import SpeechToTextResponse
@@ -35,6 +36,7 @@ class RawSpeechToTextClient:
         *,
         file: core.File,
         model: typing.Optional[SpeechToTextModel] = OMIT,
+        mode: typing.Optional[Mode] = OMIT,
         language_code: typing.Optional[SpeechToTextLanguage] = OMIT,
         input_audio_codec: typing.Optional[InputAudioCodec] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
@@ -61,12 +63,49 @@ class RawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2.5`
+
+            - **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.**
+
+            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.
@@ -85,6 +124,7 @@ class RawSpeechToTextClient:
             method="POST",
             data={
                 "model": model,
+                "mode": mode,
                 "language_code": language_code,
                 "input_audio_codec": input_audio_codec,
             },
@@ -209,7 +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.
@@ -329,6 +372,7 @@ class AsyncRawSpeechToTextClient:
         *,
         file: core.File,
         model: typing.Optional[SpeechToTextModel] = OMIT,
+        mode: typing.Optional[Mode] = OMIT,
         language_code: typing.Optional[SpeechToTextLanguage] = OMIT,
         input_audio_codec: typing.Optional[InputAudioCodec] = OMIT,
         request_options: typing.Optional[RequestOptions] = None,
@@ -355,12 +399,49 @@ class AsyncRawSpeechToTextClient:
 
         model : typing.Optional[SpeechToTextModel]
             Specifies the model to use for speech-to-text conversion.
-            Note:- Default model is `saarika:v2.5`
+
+            - **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.**
+
+            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.
@@ -379,6 +460,7 @@ class AsyncRawSpeechToTextClient:
             method="POST",
             data={
                 "model": model,
+                "mode": mode,
                 "language_code": language_code,
                 "input_audio_codec": input_audio_codec,
             },
@@ -503,7 +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_job/client.py

@@ -12,6 +12,7 @@ from ..types.files_upload_response import FilesUploadResponse
 from ..types.job_status_v_1_response import JobStatusV1Response
 from ..types.speech_to_text_model import SpeechToTextModel
 from ..types.speech_to_text_language import SpeechToTextLanguage
+from ..types.mode import Mode
 from .raw_client import AsyncRawSpeechToTextJobClient, RawSpeechToTextJobClient
 from .job import AsyncSpeechToTextJob, SpeechToTextJob
 
@@ -72,7 +73,9 @@ class SpeechToTextJobClient:
         )
         """
         _response = self._raw_client.initialise(
-            job_parameters=job_parameters, callback=callback, request_options=request_options
+            job_parameters=job_parameters,
+            callback=callback,
+            request_options=request_options,
         )
         return _response.data
 
@@ -145,11 +148,17 @@ class SpeechToTextJobClient:
             job_id="job_id",
         )
         """
-        _response = self._raw_client.start(job_id, ptu_id=ptu_id, request_options=request_options)
+        _response = self._raw_client.start(
+            job_id, ptu_id=ptu_id, request_options=request_options
+        )
         return _response.data
 
     def get_upload_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesUploadResponse:
         """
         Start a speech to text bulk job V1
@@ -180,11 +189,17 @@ class SpeechToTextJobClient:
             files=["files"],
         )
         """
-        _response = self._raw_client.get_upload_links(job_id=job_id, files=files, request_options=request_options)
+        _response = self._raw_client.get_upload_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     def get_download_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesDownloadResponse:
         """
         Start a speech to text bulk job V1
@@ -215,12 +230,15 @@ class SpeechToTextJobClient:
             files=["files"],
         )
         """
-        _response = self._raw_client.get_download_links(job_id=job_id, files=files, request_options=request_options)
+        _response = self._raw_client.get_download_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     def create_job(
         self,
         model: SpeechToTextModel = "saarika:v2.5",
+        mode: typing.Optional[Mode] = None,
         with_diarization: bool = False,
         with_timestamps: bool = False,
         language_code: typing.Optional[SpeechToTextLanguage] = None,
@@ -236,6 +254,10 @@ class SpeechToTextJobClient:
         model : SpeechToTextModel, default="saarika:v2.5"
             The model to use for transcription.
 
+        mode : typing.Optional[Mode], default=None
+            Mode of operation. Only applicable for saaras:v3 model.
+            Options: transcribe, translate, indic-en, verbatim, translit, codemix
+
         with_diarization : typing.Optional[bool], default=False
             Whether to enable speaker diarization (distinguishing who said what).
 
@@ -244,7 +266,7 @@ class SpeechToTextJobClient:
 
         language_code : typing.Optional[SpeechToTextLanguage], default=None
             The language code of the input audio (e.g., "hi-IN", "bn-IN").
-            
+
         num_speakers : typing.Optional[int], default=None
             The number of distinct speakers in the audio, if known.
 
@@ -263,6 +285,7 @@ class SpeechToTextJobClient:
             job_parameters=SpeechToTextJobParametersParams(
                 language_code=language_code,
                 model=model,
+                mode=mode,  # type: ignore[typeddict-item]
                 num_speakers=num_speakers,  # type: ignore[typeddict-item]
                 with_diarization=with_diarization,
                 with_timestamps=with_timestamps,
@@ -350,7 +373,9 @@ class AsyncSpeechToTextJobClient:
         asyncio.run(main())
         """
         _response = await self._raw_client.initialise(
-            job_parameters=job_parameters, callback=callback, request_options=request_options
+            job_parameters=job_parameters,
+            callback=callback,
+            request_options=request_options,
         )
         return _response.data
 
@@ -392,7 +417,9 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.get_status(job_id, request_options=request_options)
+        _response = await self._raw_client.get_status(
+            job_id, request_options=request_options
+        )
         return _response.data
 
     async def start(
@@ -439,11 +466,17 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.start(job_id, ptu_id=ptu_id, request_options=request_options)
+        _response = await self._raw_client.start(
+            job_id, ptu_id=ptu_id, request_options=request_options
+        )
         return _response.data
 
     async def get_upload_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesUploadResponse:
         """
         Start a speech to text bulk job V1
@@ -482,11 +515,17 @@ class AsyncSpeechToTextJobClient:
 
         asyncio.run(main())
         """
-        _response = await self._raw_client.get_upload_links(job_id=job_id, files=files, request_options=request_options)
+        _response = await self._raw_client.get_upload_links(
+            job_id=job_id, files=files, request_options=request_options
+        )
         return _response.data
 
     async def get_download_links(
-        self, *, job_id: str, files: typing.Sequence[str], request_options: typing.Optional[RequestOptions] = None
+        self,
+        *,
+        job_id: str,
+        files: typing.Sequence[str],
+        request_options: typing.Optional[RequestOptions] = None,
     ) -> FilesDownloadResponse:
         """
         Start a speech to text bulk job V1
@@ -533,6 +572,7 @@ class AsyncSpeechToTextJobClient:
     async def create_job(
         self,
         model: SpeechToTextModel = "saarika:v2.5",
+        mode: typing.Optional[Mode] = None,
         with_diarization: bool = False,
         with_timestamps: bool = False,
         language_code: typing.Optional[SpeechToTextLanguage] = None,
@@ -548,6 +588,10 @@ class AsyncSpeechToTextJobClient:
         model : SpeechToTextModel, default="saarika:v2.5"
             The model to use for transcription.
 
+        mode : typing.Optional[Mode], default=None
+            Mode of operation. Only applicable for saaras:v3 model.
+            Options: transcribe, translate, indic-en, verbatim, translit, codemix
+
         with_diarization : typing.Optional[bool], default=False
             Whether to enable speaker diarization (distinguishing who said what).
 
@@ -556,8 +600,8 @@ class AsyncSpeechToTextJobClient:
 
         language_code : typing.Optional[SpeechToTextLanguage], default=None
             The language code of the input audio (e.g., "hi-IN", "bn-IN").
-            
-        num_speakers : typing.Optional[int], default=None
+
+        num_speakers : typing.Optional[int] = None
             The number of distinct speakers in the audio, if known.
 
         callback : typing.Optional[BulkJobCallbackParams], default=OMIT
@@ -575,6 +619,7 @@ class AsyncSpeechToTextJobClient:
             job_parameters=SpeechToTextJobParametersParams(
                 language_code=language_code,
                 model=model,
+                mode=mode,  # type: ignore[typeddict-item]
                 with_diarization=with_diarization,
                 with_timestamps=with_timestamps,
                 num_speakers=num_speakers,  # type: ignore[typeddict-item]

sarvamai/speech_to_text_streaming/__init__.py

@@ -7,6 +7,8 @@ from .types import (
     SpeechToTextStreamingHighVadSensitivity,
     SpeechToTextStreamingInputAudioCodec,
     SpeechToTextStreamingLanguageCode,
+    SpeechToTextStreamingMode,
+    SpeechToTextStreamingModel,
     SpeechToTextStreamingVadSignals,
 )
 
@@ -15,5 +17,7 @@ __all__ = [
     "SpeechToTextStreamingHighVadSensitivity",
     "SpeechToTextStreamingInputAudioCodec",
     "SpeechToTextStreamingLanguageCode",
+    "SpeechToTextStreamingMode",
+    "SpeechToTextStreamingModel",
     "SpeechToTextStreamingVadSignals",
 ]