Velora
Star

Private voice typing for Mac

Say the messy version.
It types the clean one.

Hold a key and talk the way you actually talk: ums, pauses, “wait, no, Thursday.” Velora turns it into clean text in whatever app you’re in. The models that hear you and tidy you up live on your Mac, so after the one-time setup, dictation works in airplane mode.

Download for Mac Install with Homebrew
  • Free & open source
  • No account. No Velora server.
  • Apple Silicon · macOS 14+
Messages Message mode

The um is gone. The “actually” got applied. Meaning untouched.

Heard: “tell Maya um I can do three, actually make that three thirty”

Tell Maya I can do 3:30.

Ready 0.0s

Works wherever your cursor is blinking.

  • Mail
  • Slack
  • Notes
  • VS Code
  • Chrome
  • Messages
  • Notion
  • Terminal
  • Obsidian
  • Gmail
  • WhatsApp
  • Xcode
  • Pages
  • Linear
  • Cursor
  • Safari
  • Discord
  • Figma

The mechanic

One key. Zero commands to memorise.

You never say “comma” or “new line.” You hold a key, you talk, you let go. Everything between those two moments happens on your Mac.

  1. Hold

    Right Option by default, or any key you want. A small pill shows up near your cursor and never steals focus from what you’re typing into. Close the Velora window and the pill stays; hide it from its right-click menu or the menubar.

  2. Talk

    Ramble. Pause. Change your mind halfway through a sentence. Velora is built for the way people actually speak, not the way dictation software wishes they did.

  3. Release

    Cleanup runs on your Mac and the text lands at your cursor. It’s on your clipboard too, so if an app refuses the paste, ⌘V finishes the job.

Destination-aware

The app you’re in decides the format.

You don’t talk to your group chat the way you email your landlord, and Velora knows the difference. It checks which app is in front (and in Safari, Chrome or Firefox, which site) and switches its rules. Slack gets a message. Gmail gets paragraphs. Notion gets a note. Your terminal gets the command back exactly as you said it, with no trailing full stop to break it.

Message

Slack · Messages · WhatsApp

“are we still on for thursday i can move things around if not”

Are we still on for Thursday? I can move things around if not

Terse. No invented “Hey!” A short line keeps its missing full stop.

Email

Mail · Gmail · Outlook

“thanks for sending that over i read it last night separately we should talk about the pricing page before friday”

Thanks for sending that over — I read it last night.

Separately, we should talk about the pricing page before Friday.

Paragraph breaks where the topic shifts. Never invents a sign-off you didn’t say.

Note

Notes · Notion · Obsidian

“first pull the logs second check the retry counter third file the ticket”

1. Pull the logs
2. Check the retry counter
3. File the ticket

Spoken ordinals become a numbered list. Plain prose stays prose.

Code

VS Code · Cursor · Terminal

“git commit dash m fix the retry backoff”

git commit -m "fix the retry backoff"

Verbatim. Nothing capitalised, no trailing period, because a period breaks a shell command.

Beyond dictation

The parts you only notice when they’re missing.

Stream Typing

Watch it type. Fix it while you talk.

Press ⌃⇧S and Velora shows a live draft as you speak, follows corrections like “3 p.m., no, 6 p.m.,” and finishes with clean text.

Editors update at your cursor. Terminals show the draft in Velora, then insert it when you finish.

How Stream Typing works

Action Mode

Say the errand. Keep doing your thing.

Hold ⌃⇧A and say “message Priya on Slack that I’m running late.” Velora plans the steps on your Mac and carries them out in the app you’re in: a few at a time, reading what the window actually says, stopping instead of guessing. Install the open-source Cua Driver and it can drive another app in a background window while you stay in yours. Say “draft” and it writes the message but doesn’t send it.

How Action Mode works

New Proofread

Fix the typos. Keep your voice.

Select any text, press ⌃⇧G. Spelling, punctuation and grammar get fixed in place; your phrasing and the names in your dictionary are left alone. No microphone involved.

How Proofread works

Voice Edit

Select it. Say what to change.

Highlight a sentence, hold ⌥⇧E, and describe the edit out loud. It’s rewritten in place, and ⌘Z puts it back exactly as it was.

How Voice Edit works

Meeting notes

Records the room. Never the screen.

After you confirm, every time, Velora captures microphone and computer audio, keeps a visible recording indicator running, and writes a transcript and notes you can search on your Mac. Audio only.

How meeting notes work

AirPods, unharmed

Your music pauses itself.

For direct dictation, Velora pauses supported music players and resumes only what it paused, so your AirPods don’t drop into crunchy call-quality audio mid-song. Browser video is excluded from automatic pause and resume. Meeting capture leaves call audio running.

History

Nothing is thrown away.

Every dictation keeps its raw transcript next to the cleaned version, searchable in History, on your Mac. If cleanup got it wrong, the original is one click away.

How history works

Languages

99 languages, one model.

The default speech model is Whisper large-v3-turbo: multilingual, including Hindi and Indian English. Set a language or let Velora detect it.

Languages & models

Personal dictionary

It learns your nouns.

Names, products and jargon it got wrong once are learned from your corrections and from words that keep showing up. Sync runs through your own iCloud Drive whenever it is signed in, never through a Velora server.

How the dictionary works

Stats

See what you saved.

Words dictated, speaking time, a 12-week streak heatmap and the apps you talk into most. Computed on your Mac, shared only if you export a card.

Everything, indexed

Every feature has its own page.

No mystery-meat marketing. Each feature gets a page with the exact shortcut, what it does, and the fine print on what it won’t.

Hold-to-talk dictation

Hold Right Option, talk, let go. Clean text lands at your cursor in any Mac app.

Stream Typing

⌃⇧S shows a live draft while you speak and follows your mid-sentence corrections.

Action Mode

⌃⇧A turns a spoken sentence into real steps in real apps, planned and run on your Mac.

ProofreadNew

⌃⇧G fixes spelling, punctuation and grammar in whatever you selected. No microphone.

Voice Edit

Select text, say the change, get it rewritten in place, with ⌘Z as the exit.

Meeting notes

Consent-first capture of mic and system audio, with local transcripts and notes.

App-aware modes

Slack gets a message, Gmail gets paragraphs, your terminal gets verbatim commands.

Personal dictionary

Your names, products and jargon, learned from your corrections, synced through your own iCloud Drive whenever it is signed in.

History & recovery

Raw transcript and final text for every dictation, searchable and replayable on your Mac.

Languages & models

99 languages through Whisper, faster English-first options through Parakeet.

File transcription

Open audio or video with Velora (MP3, WAV, MOV, MP4) and get a local transcript.

The privacy model

What leaves your Mac and when, traced to the code rather than a policy page.

Private by architecture

Your voice has nowhere to go.

Most dictation apps promise privacy in a document you have to trust. Velora’s speech model and cleanup model are files on your disk, so there is no Velora server for your audio to travel to. Turn on airplane mode after setup and dictation keeps working, because the whole thing runs on your Mac. Three things do use the network, and the ledger names them.

Read the privacy model
Your voice
stays on the Mac
Transcripts & history
stay on the Mac
Meeting audio & notes
stay on the Mac
Screen context (window titles, on-screen text, optional local text recognition)
stays on the Mac
Accounts, telemetry, ads
do not exist
Model download, first run (and first use of speaker separation)
uses the network once each
Update checks & installs
GitHub, optional, can be turned off
Dictionary sync
your iCloud Drive, whenever it is signed in

The honest comparison

Same job. Very different deals.

Dictation tools differ on three things: where your audio goes, what the polish costs, and whether you can check. The table below covers all of it, including what the others do better.

Checked September 2026 · full detail on the comparison pages
Velora Wispr Flow Superwhisper Apple Dictation
Price Free forever, MIT-licensed Free tier capped at 2,000 words/week; Pro $12–15/mo Free tier; Pro $8.49/mo or $249.99 lifetime Free, built in
Where your audio goes Stays on your Mac Their servers — “transcription always occurs on the cloud” On-device with local models; optional cloud models On-device for supported languages
Account required Never Yes No account; licence unlocks Pro Apple ID is part of macOS
AI cleanup of what you said Local LLM, included free In the cloud Whisper models free; other AI models are Pro or your own API key None — raw transcription
Open source All of it, on GitHub No No No
Spoken actions in real apps Action Mode, local No No No
Meeting notes Local, consent-first Cloud notetaker Yes No
Platforms Mac (Apple Silicon) Mac, Windows, iPhone, Android Mac, Windows, iPhone, Android Every Apple device

The cloud quote is from Wispr Flow’s own data-controls page. Prices and claims checked September 2026; corrections welcome — open an issue.

Velora vs Wispr Flow

Beautiful product, cloud architecture. Why “privacy mode” is not the same as your audio never leaving.

Read the comparison

Velora vs Superwhisper

The closest competitor, and also local-capable. The differences are the licence, the cleanup, and the price.

Read the comparison

The whole field

Wispr Flow, Superwhisper, Apple Dictation, and open-source FluidVoice, one honest matrix.

See every tool

Alpha Velora for iPhone

An early experiment, not a product yet.

The iPhone companion turns the Action Button into a voice key: press, speak, finish, and the transcript is on your clipboard. Recognition is on-device, using Apple’s own speech models.

It is alpha and it is not stable. Expect rough edges, bugs, and breaking changes between commits. It is not on the App Store, there is no TestFlight build, and it is not something to depend on. Build it only if you are comfortable running unfinished software from source.

Build it from source Action Button setup

Install

Two minutes to your first sentence.

Download the signed, notarised app or install it with Homebrew. On first launch Velora asks for microphone, input-monitoring and accessibility permissions, then downloads the models it needs, once. After that, dictation works in airplane mode.

Download for Mac Star on GitHub
Terminal 
$ brew install --cask sushilk1991/tap/velora

Velora is ready in Applications

Straight answers

Before you install.

Does Velora ever touch the network?

Three ways, and never with your voice. The model download happens once, during setup, from Hugging Face, plus a 46 MB speaker-separation download from GitHub the first time you use meeting notes. Optional update checks ask GitHub for the latest release, and optional automatic installs download the new build from GitHub; both can be turned off in Settings. Dictionary sync goes through your own iCloud Drive whenever it is signed in and carries confirmed terms only; there is no Velora server. During dictation, speech recognition and cleanup run on your Mac; your audio and transcript are not uploaded.

Does it work offline?

Yes. Once the models are downloaded, dictation, Stream Typing, Voice Edit, Proofread and meeting notes all work with no connection at all. Airplane mode is a fine way to check.

What Mac do I need?

An Apple Silicon Mac running macOS 14 or later. Disk use depends on which speech and cleanup models you pick; Velora chooses a cleanup tier based on how much RAM you have.

Is it really free?

Yes. It’s MIT-licensed and open source, with no account, no subscription and no paid tier. There is nothing to upsell you to because there is no server to pay for.

Which apps does it work in?

Any Mac app with a text field. Velora pastes at your cursor, so Mail, Slack, Notes, VS Code, browsers and terminals all work. If an app refuses the paste, the text is on your clipboard.

Why not just use Apple’s built-in dictation?

Apple Dictation types what it hears. Velora also cleans it up: filler out, “wait, no” applied, formatted for the app you’re in, your names spelled the way you spell them. And it keeps the raw transcript in History in case the cleanup gets clever.

Is the iPhone app in the App Store?

No — and it is worth being blunt: the iPhone app is alpha and unstable. The Xcode project is in the repository and builds on iOS 17 or later, but it is unfinished, changes without warning, and is not something to rely on. The Mac app is the real product.

How do I get help?

Email sushilk.1991@gmail.com or open a GitHub issue. Velora is pre-1.0 and actively developed; bug reports genuinely help.

Free. Open source. Local.

Stop typing. Keep writing.

Get Velora for Mac Star on GitHub

Already using Velora? A star on GitHub is the entire marketing budget.