OpenTypeless

SiliconFlow STT

SiliconFlow STT uses SenseVoiceSmall through an OpenAI-compatible transcription endpoint.

Official Documentation

SiliconFlow Audio Transcription API

Official transcription endpoint and request fields.

https://docs.siliconflow.com/cn/api-reference/audio/createTranscription

SiliconFlow API Key Console

Create and manage access keys.

https://cloud.siliconflow.cn/account/ak

Get API Key

  1. Sign in to SiliconFlow cloud console.
  2. Create access key in AK page.
  3. Store key securely.

Configure in OpenTypeless

  1. Settings > Speech Recognition > Provider: SiliconFlow.
  2. Paste key and run Test.
  3. Use fixed language only when workflow is single-language and latency-sensitive.

Recommended Configuration

OpenTypeless uses SenseVoiceSmall on SiliconFlow transcription endpoint.

Provider: siliconflow (STT)
Protocol: HTTPS multipart upload
Endpoint: https://api.siliconflow.cn/v1/audio/transcriptions
Auth header: Authorization: Bearer <SILICONFLOW_API_KEY>
Model: FunAudioLLM/SenseVoiceSmall

Operational Tip

💡
Run short benchmark loops against Groq/Deepgram before deciding your default provider.