Yandex SpeechKit
Synthesize speech and recognize audio via Yandex SpeechKit
ya
Yandex SpeechKit is Yandex Cloud's speech technology platform offering text-to-speech (TTS) synthesis and speech-to-text (STT) recognition for Russian and other languages.
With the Yandex SpeechKit integration in AACFlow, you can:
- Text-to-Speech (Synthesize): Convert text to audio in various voices and formats
- Speech-to-Text (Recognize Short): Transcribe short audio files (under 1 minute) to text
This integration enables automated voice response systems, audio content generation, and speech-driven workflow triggers.
Usage Instructions
Integrate Yandex SpeechKit into the workflow to add voice synthesis or audio transcription. Requires a Yandex Cloud IAM token. Obtain IAM tokens using the Yandex Cloud IAM block in your workflow.
Tools
yandex_speechkit_tts
Text-to-speech synthesis
Input
| Parameter | Type | Required | Description |
|---|---|---|---|
iamToken | string | Yes | Yandex Cloud IAM token |
text | string | Yes | Text to convert to speech |
voice | string | No | Voice name (e.g., oksana, alena, filipp) |
speed | number | No | Speech speed (0.1–3.0, default 1.0) |
format | string | No | Audio format: ogg_opus, lpcm, mp3 |
Output
| Parameter | Type | Description |
|---|---|---|
audioData | string | Base64-encoded audio data |
mimeType | string | Audio MIME type |
yandex_speechkit_stt
Speech-to-text recognition (short audio)
Input
| Parameter | Type | Required | Description |
|---|---|---|---|
iamToken | string | Yes | Yandex Cloud IAM token |
audioData | string | Yes | Base64-encoded audio data |
language | string | No | Language code (ru-RU, en-US) |
Output
| Parameter | Type | Description |
|---|---|---|
text | string | Recognized text |
confidence | number | Recognition confidence score |
Links
- Official API: https://yandex.cloud/ru/docs/speechkit/
- AACFlow block source: apps/aacflow/blocks/blocks/yandex_speechkit.ts