Dettivo

§ STT engines

STT engines

Dictation and meetings can pick different STT models. Whisper tiers (via the WhisperKit engine) and Parakeet (via the FluidAudio engine) ship by default, both run on-device on Apple Silicon. Cloud STT is opt-in, BYOK, and currently used for dictation paths only.

Who this is for

Users picking the right speed/accuracy tradeoff for their Mac, and anyone who wants to run hosted STT for specific recordings.

Choosing a tier

Smaller Whisper tiers run faster with lower accuracy on long-form audio. Larger tiers are slower but cleaner. On an M1 Max or better, the larger tiers are usually fast enough for live transcription. On an M1 Air, the smaller tiers (or Parakeet) are the better default.

Where STT engines are configured

Settings › Models

Configuration target
Pick whether you are configuring Dictation or Meetings.
Local engine
Whisper tier picker (via the WhisperKit engine) and the Parakeet model (via FluidAudio). Download or remove per model.
Cloud Providers (BYOK)
OpenAI Whisper, Anthropic, OpenAI-compatible endpoints. Dictation-oriented.
Model stats
WER, best use case, language support, visible per tier.
§ Related

Just launched

Just launched on your Mac.

One-time license, three active Macs, lifetime updates. $45 once, forever yours.