Private, on-device dictation for Mac and iPhone

Say it loud. Keep it private.

Professional voice dictation, made easy.
On your device — private. No cloud, no uploads.

On Mac, hold a hotkey. On iPhone, hold the mic or use the Outspoken keyboard. Your voice becomes clean text locally, wherever you write.

Download for macOS Available now
iPhone beta · in TestFlight

macOS 14+ · Apple silicon · iPhone on iOS 17+ · no account

  • on-device transcription
  • works offline
  • Swiss German beta
  • made in Switzerland

raw in · polished out

Say it messy.
It lands clean.

Fillers dropped. Grammar fixed. Meaning kept — by your local model.

01The flow

Press, speak, release.

No app to open, no window to switch to. Outspoken lives behind a hotkey and types where your cursor already is.

1
right ⌘

Hold the hotkey

Anywhere, in any app. A small overlay appears at the bottom of your screen.

2

Speak naturally

Ums, false starts, mid-sentence pivots — say it the way you'd say it out loud.

3
~0.1–0.3 s to text with the model warm (Small · measured on M-series) · AI rewrite +0.3–0.5 s (optional)

Release

whisper.cpp transcribes on-device. Your local model tidies grammar, punctuation, and formatting.

4
Could we move the retro

Text lands in place

Polished text lands at your cursor in one atomic insert — and stays one ⌘V away as a fallback.

02Privacy model

Local by design, not by settings toggle.

There is no Outspoken cloud. Audio is captured and transcribed in-process. If you enable rewriting, the transcript goes to a model server on localhost — anything that isn't 127.0.0.1 is refused outright, redirects included, unless you explicitly approve your own server: one host, spelled out by you, off by default. The only other network access: Whisper model downloads you start yourself, checksum-verified, from huggingface.co. Every connection is listed in Settings → Network activity.

Where your data actually lives

audio
Processed in memory. Never written to disk, never uploaded.
transcripts
Saved only if you turn on History — plain local files under ~/Library/Application Support/Outspoken/.
dictionary & snippets
One local JSON file. Edit it by hand, version it, back it up.
rewrite prompts
Sent only to the endpoint you configure. Default: nowhere.
account
None. Download, drag to Applications, talk.

03Every app

If you can type there, you can speak there.

Outspoken types wherever your cursor is — native apps, Electron apps, browser text fields — and matches the register of the room: full sentences in email, loose in chat, a tidy comment in your editor.

Mail

Hi Priya — the revised SOW is attached, with the timeline moved up a week. Could you review section 3 before Friday?

tone · professional
Slack

shipping the fix now — should be live in ~10 min. will post in #incidents when it lands

tone · casual
Notion

Decision: consolidate both onboarding flows before the October release. Owner: Dana. Revisit at the next planning review.

structure · notes & docs
Cursor

// Debounce resize events so the chart doesn't re-render on every frame

context · code comment
Linear

Date picker resets to UTC when the browser locale is set to JST

format · issue title
Safari

coffee roasters near bernal heights open before 7am

anywhere · any text field
  • Macavailable now
  • iPhonein TestFlight

One voice. Two Apple platforms.

On Mac, Outspoken types into the app in front of you. On iPhone, dictate in Outspoken or use its keyboard in Messages, Mail and wherever you write. Transcription stays on the device on both.

  • Dictate across appsThe Outspoken keyboard carries your on-device dictation into the text field you are using.
  • Swiss German betaAvailable as a real language profile on Mac and iPhone.
  • Live interpreterHave a two-way translated conversation directly on your iPhone.
Download for macOS Available now
iPhone beta · in TestFlight

04Built-in local AI

Local AI, built right in.

Outspoken cleans up grammar, removes fillers, formats and restyles your words directly on your Mac — no Ollama, LM Studio or separate server setup required. The engine ships with the app; download a model once, then keep everything local. Already run your own AI? Connect that instead. Prefer verbatim? Switch rewriting off.

Outspoken · Settings

Transcription

EngineBundled, runs in-process
whisper.cpp
ModelCPU decode via Apple's Accelerate framework
small · 488 MB

Rewriting

ProviderCleans grammar, punctuation & tone
OffBuilt-inOllamaCustom
EngineShips with Outspoken, runs only on this Mac
llama.cpp · built in
Model
Qwen2.5 7B · local
ready · on-device

Network

Non-local endpointsRefused by default. Your own server needs an explicit consent toggle — and is limited to exactly that one host.
refused by default
  • No separate AI app

    The local engine ships with Outspoken. Download the model you want once, then work offline.

  • Models are files, not subscriptions

    Whisper model files (GGML) on your disk. Swap or delete them anytime — downloads happen in-app, SHA-256-verified.

  • Raw mode

    Turn rewriting off for verbatim transcripts — quotes, minutes, legal.

05The details

Built for people who dictate all day.

Custom dictionary

Teach it “Kubernetes”, “Priya”, “PostgREST” once. Lives with your snippets in one local JSON file you can edit by hand — changes apply instantly.

Voice snippets

Say “insert my address” and the whole block lands — sign-offs, links, canned replies. Snippet contents never reach the AI server.

Formatting commands

“New paragraph.” “Bullet point.” “Question mark.” Structure and punctuation are spoken, not typed — in English and German.

App-aware tone

Mail comes out professional, Slack stays casual, your editor gets a code comment — picked from the app in front. One toggle turns it off.

Raw transcript mode

One toggle for verbatim output — every word as said, no model in between.

Swiss German, in beta

Speak Swiss German. Outspoken understands it and can keep the result in dialect — with local models and regional controls in Settings.

Live translation

Dictate in German, English lands. Other target languages (French, Spanish, Italian, …) use your local AI model — all on this Mac.

Secure fields stay off-limits

When macOS flags secure input — password fields, login prompts — nothing is typed or pasted. Ever.

Clipboard-safe insertion

Text lands in one atomic insert — nothing dropped, nothing doubled. A clipboard rescue keeps every dictation one ⌘V away if an app refuses the insert.

06Choose your platform

Outspoken for Mac and iPhone.

The same private foundation, shaped for how each Apple platform works. Get the Mac beta today; the iPhone beta is already in TestFlight.

macOS Available now

Download the Mac beta.

Outspoken 0.4.0 runs on macOS 14+ and Apple silicon. Hold a global hotkey and dictate into Mail, Slack, your editor or any browser field.

  • Built-in multilingual Whisper model
  • Local AI cleanup and formatting
  • Signed and notarized by Apple
Download for macOS

Outspoken-0.4.0.dmg · ~91 MB · no account

iPhone In TestFlight

Outspoken for iPhone.

The iPhone app is already being tested through TestFlight. Dictate in Outspoken or use its keyboard across apps, with transcription and translation running on the device.

  • On-device dictation across apps
  • Swiss German beta
  • Live interpreter on iPhone

iOS 17+ · TestFlight beta · public release next

07Questions

Fair questions.

How is this different from Wispr Flow?

Cloud dictation tools like Wispr Flow send your audio to their servers for transcription and cleanup. Outspoken does the same job — hotkey, polish, every app — with models that run on your machine. No audio, text, or context ever reaches us, because there's no “us” in the loop at all.

Does it work offline?

Yes, on Mac and iPhone. Transcription runs in-process on the device; optional models stay local once downloaded. Network access is limited to model downloads you start — and, on Mac, a server you explicitly choose to connect.

Where does the rewrite step run?

By default, on Outspoken's built-in local engine. If you choose your own endpoint instead, the transcript plus your style rules go exactly there. Snippet contents are swapped for placeholders before the request, so they never leave the app. Turn rewriting off and no request is made at all.

Which models does it use?

Speech-to-text runs on the whisper.cpp family, from tiny to large-v3 — including large-v3-turbo and a quantized q5 build. For rewriting, choose one of Outspoken's local models or connect any chat model behind an OpenAI-compatible API. Downloads happen in the app and are SHA-256-verified.

Can it translate while I dictate?

Yes. On Mac, speak German and English comes out; other target languages use your local AI model. On iPhone, local translation profiles and a live interpreter cover supported language pairs without sending the conversation to us.

What does the Mac beta require?

Any Apple silicon Mac running macOS 14 or later. The default small model (~490 MB) decodes well over 30× realtime in our M-series measurements; lighter and larger models are one click away in Settings.

Is Outspoken available on iPhone?

Yes — the iPhone app is currently in TestFlight on iOS 17 and later. Its public App Store release is coming soon; the macOS beta is available to download today.

What about Windows or Linux?

Outspoken is focused on Mac and iPhone today. Windows and Linux remain on the roadmap.

Mac + iPhone

Talk more. Type less.

Outspoken for Mac is ready to download. Outspoken for iPhone is already in TestFlight and coming soon to the App Store.

Download for macOS Available now
iPhone beta · in TestFlight

macOS 14+ · Apple silicon · iOS 17+ · no account