The voice interface
for your computer.

Dictate anywhere. Capture every meeting. Build voice workflows.
Fast and on-device.

FinderFileEditViewWindowHelp
25:00
A
Home
DMs
Activity
More
Y
Slack
Unreads
Drafts and sent
Channels
# design
# engineering
Direct messages
S
Sam1
J
Jordan2
#engineering
2
👋  Welcome to the #engineering channel
This channel is for the engineering team. Share updates, discuss architecture, and coordinate on shipping great software.
Today
S
Sam9:47 AM
Auth module is throwing token expiry errors in prod — can someone take a look? 🔴
Message #engineering
Aa
Today
New note
No addition...
Dream where I fo...
Jan 28When I put it t...
List of books
Jan 15Doctor Faustu...
client.ts
src api client.ts
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
export async function fetchUser(id: string)
const res = await fetch(`/api/users/${id}`)
if (!res.ok) throw new Error('Failed')
return res.json()
}
 
export async function createProject(data: ProjectInput)
const headers = getAuthHeaders()
const body = JSON.stringify(data)
const res = await fetch('/api/projects', {
method: 'POST', headers, body
})
if (!res.ok) throw new ApiError(res)
return res.json()
}
 
// Retry wrapper for transient failures
export async function withRetry<T>(
fn: () => Promise<T>,
retries = 3
): Promise<T>
// TODO: implement backoff
The token refresh is handled in refreshToken() at line 42. The current catch block silently swallows errors. Here's how to add retry logic with exponential backoff:
async function refreshToken(retries = 3) {
for (let i = 0; i < retries; i++) {
try { return await fetchNewToken() }
catch { await sleep(2 ** i * 1000) }
}
}
auth.tsCurrent File
Downloads
Favorites
Desktop
Documents
Downloads
NameDate ModifiedSizeKind
banner.webpToday, 9:14 AM2.3 MBWebP Image
photo.jpgToday, 8:30 AM4.2 MBJPEG Image
PDFreport-q1.pdfYesterday1.8 MBPDF Document
PDFreport-q2.pdfYesterday2.1 MBPDF Document
config.jsonFeb 4, 202612 KBJSON File
notes.txtFeb 3, 20264 KBText File
Finder
Finder
Cursor
Cursor
Slack
Slack
Notes
Notes

Trusted by thousands of early adopters building with voice.

75M+
Words Spoken
Unlimited Dictation
50K+
Hours Saved
Offline First
20K+
Meetings transcribed

Three layers.
Built around your choices.

Cloud
Groq
Groq
Whisper Large v3 Turbo
Soniox
Soniox
Real-Time STT
Setup
Local
Parakeet V2
Parakeet V2
430 MB
Small
Small
465 MB

Your speech-to-text

Local by default — your audio stays on your Mac. Cloud when you want more accuracy.

GROQ

Your AI

Bring your own provider key. Cleans up dictation, summarizes meetings, turns voice into action.

fn⌘ ⇧

Your shortcut

Press a key. Speak. Get back what you need — polished text in any app, a meeting summary, or a workflow that runs the action.

Voice tools that just work.
Wherever you are.

Your words, expanded effortlessly.

Speak the shortcut and Vowen writes the full message — out-of-office notes, intros, elevator pitches, anything.

my calendarhttps://calendly.com/dominic
ooo responseI'll be out of office until next Monday — for urgent issues...
introHi, I'm Dominic — Lead Engineer at Vowen.
elevator pitchVowen is the voice interface for your computer.

Private by designAll your data stays yours.

100+ languagesand dialects, transcribed natively. Real-time translation to English, always on.

helloவணக்கம்ciaoприветoláহ্যালো안녕하세요bonjouralohaγειάمرحبا你好नमस्तेשלוםxin chàoholaguten tagสวัสดีahojterehujambokia orahejdzień dobryسلامmoigod dagmerhaba
EnglishSpanishFrenchGermanJapaneseMandarinEnglishSpanishFrenchGermanJapaneseMandarin
RussianPortugueseDutchPolishTurkishItalianRussianPortugueseDutchPolishTurkishItalian

Custom Instructions — per app, or one for everything.

Tell Vowen how to sound like you.

“Be concise. Sign off with my name. Never use emojis.”

Slack
Casual
Email
Formal
Notion
Bullets
Linear
Terse

Custom Instructions — per app, or one for everything.

Tell Vowen how to sound like you.

“Be concise. Sign off with my name. Never use emojis.”

Slack
Casual
Email
Formal
Notion
Bullets
Linear
Terse

Voice to text action.
Just speak the intent.

Powered by world-class AI models

Anthropic
OpenAI
Gemini
Mistral
Groq
OpenRouter
Deepgram
ElevenLabs
AssemblyAI
Anthropic
OpenAI
Gemini
Mistral
Groq
OpenRouter
Deepgram
ElevenLabs
AssemblyAI

Meetings. Lectures. Podcasts.
Turned into notes you can ask.

Show up. We'll catch the rest.

Ask anything

Chat with any meeting. Surface decisions, quotes, or follow-ups instantly.

Auto-detect meetings

Vowen catches when you're on a call and offers to start recording. No setup.

Speaker identification

Every voice automatically labeled across the transcript. Quote anyone, instantly.

Custom templates

Define exactly how every summary should be structured and styled.

Save to Obsidian

Sync every meeting straight into your Obsidian Vault, fully formatted.

Webhooks anywhere

Post finished notes to Slack, Notion, or any URL the moment a meeting ends.

Plus PDF export, multi-model summary regeneration, and Notion-style editing. Depth where you need it.

Voice productivity beyond dictation.

Build trigger phrases. Open apps, folders, websites — run anything on your computer.

Hold Fn to trigger

Tasks that took clicks, now take words.

Native utilities built directly into the OS layer. Say the command — watch it execute instantly.

+ to trigger command mode
In action
Merge PDFs

Set Timers

hands-free focus
25:00
25:005:0015:00say “start 25 minute timer”
EN
“How are you doing today?”
ES
“¿Cómo estás hoy?”

Translate

Convert Media

.MOV.MP4
148 MB42 MB
#6366F1

Color Extractor

Compress Images

−68%
smaller, lossless

Open in Editor

Cursor
VS Code·Sublime·Zed·Xcode

Config Convert

JSON
{
  "name": "vowen",
  "port": 3000
}
YAML
name: vowen
port: 3000

Create your own
custom utility.

Assign shortcuts to trigger on demand.

Got Questions?

If you can't find what you're looking for, get in touch.

Is Vowen really free?
Yes — Vowen has a generous list of free features. Core ones include voice dictation, AI text enhancement, and custom vocabulary, among others. See everything that's free
What are the system requirements?
Vowen is available for macOS 14.0 or later on Apple Silicon (M1/M2/M3/M4) Macs, and Windows 10 or later on x64 systems. Linux version is coming soon.
How do I use Vowen?
Simply press and hold fn key to transcribe your voice, or Option+Shift for AI-powered assistance. The app runs in your menu bar and works across all your favorite apps.
Do I need to create an account?
No account needed! Download, install, and start using Vowen immediately. Your data stays completely local on your device.
Does Vowen work offline?
Yes! Vowen's transcription runs completely locally using open-source Whisper models. No internet connection required for basic transcription. AI features (optional) require internet.
Is my data private and secure?
Absolutely. Vowen does not collect, store, or access any of your voice data or text. All transcription happens locally on your device. If you choose to use AI features, only the transcribed text is processed by the AI provider.
Which AI providers are supported?
Vowen supports OpenAI (GPT-4), Anthropic (Claude), Google Gemini, Groq, DeepSeek, and OpenRouter. You can also use custom OpenAI-compatible APIs. AI features are optional and require your own API keys.
How can I request a feature or report a bug?
We're a team of two working on Vowen part-time. You can submit feature requests on Featurebase (https://vowen.featurebase.app/), join our Discord, or email us at [email protected]. We read everything.

Say it, ship it.

Try Vowen for free today. Find your flow.

macOS 14.0 or laterWindows 10 or later