On-device transcription
Speech recognition can run entirely on Apple silicon after the required local model is available.
Local AI on Apple silicon
Keep the input chain on your Mac.
Local-only mode restricts Moro to local transcription and compatible on-device models, so your content does not leave the Mac for processing.
Apple Silicon · macOS 15 Sequoia+ · Windows version in development
What it does
A clear local boundary
Local model routing
Assign compatible local models to writing, translation, chat, or autocomplete according to your workflow.
Secure fields are skipped
Moro stops in password and secure input fields instead of reading or processing their contents.
Workflow
Use local-only mode
- 01
Download the local models
Install the transcription and language models needed by the features you want to use.
- 02
Enable local-only
Moro limits supported business requests to local backends.
- 03
Check feature routing
Confirm that each enabled feature points to a compatible local model.
Common questions
A few details worth knowing
Does every feature work offline?
Core transcription, basic cleanup, notes, and supported local-model tasks can work offline. Availability depends on which local models are installed.
Is an Intel Mac supported?
Moro currently targets Apple silicon and macOS 15 or later. Local inference is optimized for arm64.
Take the shorter path from thought to text
Moro private beta is now underway. Apply to receive download access in batches.
Apply for the private beta →