VoxFlow

Speak your mind. Keep your flow.

VoxFlow turns your voice into text on your Mac. Dictate into compatible apps, transcribe audio and video, and shape your words with local AI — with reusable snippets, a personal dictionary and offline processing once your models are set up.

On-device models
Works offline
No account

Tell me when it ships Talk to us

VoxFlow welcome screen: “Speak. It types.” — VoxFlow turns your voice into clean text in any app, entirely on this Mac, with on-device models, offline use and no account

01 — DICTATE

Start with your voice

Hold a shortcut to dictate, or switch to hands-free mode for a longer thought. A compact Flow Bar shows when VoxFlow is listening and processing, and your words appear in the focused text field of compatible apps. Where direct insertion is unavailable, the finished text lands on the clipboard, ready to paste.

The VoxFlow Flow Bar while listening: a record dot, a live waveform, a 0:04 timer and a language selector

02 — STYLE

Make the result fit the moment

Use a formal tone for a considered email, a casual one for everyday writing, or Verbatim mode when the phrasing matters. Optional local AI rewrites the text while cleanup controls handle filler words, punctuation and capitalisation. Styles can differ per app, and a personal dictionary keeps names and specialist terms intact.

VoxFlow style settings: the same sentence rewritten in formal, casual and very casual tones, with per-app overrides and cleanup toggles

03 — REUSE

Keep useful phrases close

Snippets turn short triggers such as /sig into reusable text — signatures, repeatable replies, everyday templates — expanded while dictating or by typing in supported fields. Searchable history lets you revisit a dictation, copy it again or change its tone without recording. History can be encrypted on disk, or switched off.

VoxFlow snippets: triggers such as /sig, /standup and /eta expanding into signatures and repeatable replies

04 — TRANSCRIBE

Already have a recording?

Drop in an audio or video file and VoxFlow turns it into a transcript you can search and re-segment. Export as plain text, Markdown, SRT, VTT or JSON, and switch formats without transcribing the recording again — whether you need notes, subtitles, or text for another workflow.

A finished VoxFlow transcript with a search field, segment length control and TXT, SRT, VTT, JSON and MD export options

Platform

Native on macOS, quiet by default

Built for macOS on Apple Silicon, with control over your microphone, shortcuts, appearance and local history. No account required.

Tell me when it ships

Speech recognition and optional AI rewriting run locally, so you can work offline once your models are downloaded.
VoxFlow has not been released yet. Interface images come from a development build with demonstration data, and features may change before launch.