Yandex Speech
Kit integration for speech recognition Yandex SpeechKit is a Russian STT service optimized for the Russian language, Russian names, and realities. Infrastructure in the Russian Federation - data does not leave the country. WER for Russian spoken audio: 10-15%, for clear speech: 5-8%. ### Advantages for Russian-language tasks - Better quality in Russian compared to Whisper in telephony conditions - Recognition of names and toponyms of the Russian Federation without additional configuration - FSTEC compliance with on-premise deployment (SpeechKit Enterprise) - Integration with the Yandex Cloud ecosystem: Object Storage, API Gateway ### Streaming recognition via gRPC```python import grpc from yandex.cloud.ai.stt.v3 import stt_pb2, stt_pb2_grpc, stt_service_pb2
channel = grpc.secure_channel('stt.api.cloud.yandex.net:443', grpc.ssl_channel_credentials()) stub = stt_pb2_grpc.RecognizerStub(channel)
recognize_options = stt_pb2.StreamingOptions( recognition_model=stt_pb2.RecognitionModelOptions( audio_format=stt_pb2.AudioFormatOptions( raw_audio=stt_pb2.RawAudio( audio_encoding=stt_pb2.RawAudio.LINEAR16_PCM, sample_rate_hertz=16000, audio_channel_count=1 ) ), language_restriction=stt_pb2.LanguageRestrictionOptions( restriction_type=stt_pb2.LanguageRestrictionOptions.WHITELIST, language_code=['ru-RU'] ), text_normalization=stt_pb2.TextNormalizationOptions( text_normalization=stt_pb2.TextNormalizationOptions.TEXT_NORMALIZATION_ENABLED, profanity_filter=False, literature_text=True ) ) )