AssemblyAI

AssemblyAI provides robust real-time transcription with clean turn formatting.

Official Documentation

AssemblyAI Streaming Transcription

WebSocket streaming API reference and session behavior.

https://www.assemblyai.com/docs/speech-to-text/streaming

AssemblyAI Getting Started

Account setup and API key onboarding.

https://www.assemblyai.com/docs/getting-started

Get API Key

Sign in to AssemblyAI dashboard.
Generate an API key in account/project settings.
Copy and store securely.

Configure in OpenTypeless

Settings > Speech Recognition > Provider: AssemblyAI.
Paste key and run Test.
Keep default sample rate first, then tune only after baseline passes.

Recommended Configuration

OpenTypeless streams PCM chunks to AssemblyAI v3 WebSocket.

Provider: assemblyai
Protocol: WebSocket
Endpoint: wss://streaming.assemblyai.com/v3/ws
Auth header: Authorization: <ASSEMBLYAI_API_KEY>
Query params: sample_rate=16000, format_turns=true

Operational Tip

💡

If turn segmentation is unstable, test with a quieter microphone profile before switching providers.