Platform edition
AssemblyAI for macOS
Cloud transcription with AssemblyAI, real-time streaming, and REST fallback.
- Version
- 1.1.0
- Plugin ID
- com.
typewhisper. assemblyai - Requirements
- TypeWhisper 1.5.0 or newer · macOS 14.0 or newer
Inside this edition
Features in this edition
- Cloud speech-to-text processing
- Real-time transcription while recording
- Automatic REST fallback when streaming is unavailable
- Optional speaker diarization
Settings in this edition
- Store, validate, or remove the API key in macOS Keychain
- Choose the transcription model in add-on or provider settings
- Enable or disable speaker diarization; it uses the structured REST path
About this edition
The macOS edition provides Universal-3.5 Pro and Universal-2. It uses live WebSocket transcription and falls back to REST processing when the connection fails. Enabling speaker diarization selects the structured REST path and disables live preview.
Models
| Model | ID | Languages |
|---|---|---|
| Universal-3.5 Pro | universal-3-5-pro | 18 |
| Universal-2 | universal-2 | 37 |
Saved macOS selections of universal-3-pro migrate to Universal-3.5 Pro. The model and language determine the streaming endpoint. Dictionary terms are sent as model-specific recognition hints.
Setup
- Open Settings > Integrations > AssemblyAI.
- Save and validate your AssemblyAI API key.
- Choose a model and a supported language, or automatic detection.
- Enable speaker diarization only when you need it.
- Select AssemblyAI as the workflow’s transcription provider.
Audio is sent to AssemblyAI for processing. The Windows edition has its own model catalog; use the Windows guide for that edition.