Blog
Introducing SleekVoice
SleekVoice is free, open-source voice typing for macOS. Whisper runs on your Mac, and the text lands in the app you are already using.
Published · Updated · 5 min read · Bharat Bhavnasi
SleekVoice is a free, open-source voice typing app for macOS. You hold a push-to-talk shortcut, speak, and the text shows up in whichever app is focused. Transcription uses Whisper on the Mac, including when the Mac is offline. The license is GPL-3.0. The source lives at github.com/SuperAgents-Dev/SleekVoice (opens in a new tab). That is the whole product in one paragraph, and the rest of this post is what that paragraph does and does not mean.
People have been asking for a community, pull-request-driven alternative to Wispr Flow. Wispr Flow is a commercial dictation product. SleekVoice is the attempt to put that kind of voice typing in front of anyone with a Mac, without a subscription and without sending the audio to a SleekVoice server. This introduction stays inside the facts the project has published. It does not include download counts, star counts, speed claims, or quotes from users, because the project has not published any of those.
Who it is for
SleekVoice is for someone who already writes on a Mac and would rather say a paragraph than type it, in the app they are already in. That might be a note, a message, an email, a document, or an issue comment. The focused app is the destination. There is no separate transcript window you are supposed to copy out of.
It is also for someone who wants the recording processed on the machine. The project’s line is direct: your voice stays on your Mac. If you need Windows, iPhone, or Android, this is not the app. Wispr’s pricing page lists those platforms for Flow. SleekVoice does not.
It is a poor fit if you want a meeting bot, a hosted speech API, or a plugin that other software calls over the network. None of those exist for SleekVoice. The product is a Mac app you install.
What you can do with it
The published feature set is short on purpose.
Local Whisper. Speech is transcribed on the Mac. The project says this includes offline use. Whisper itself is OpenAI’s speech recognition model, released under the MIT license and documented at github.com/openai/whisper (opens in a new tab). SleekVoice has not published which model size it ships. Do not assume “large” or “tiny” until the app or the repository says so.
Push-to-talk. You hold a keyboard shortcut while you speak and release it when you are done. The project has not published the default keys. The shortcut shown in the app is the one to use.
Any focused app. The words are typed where the cursor is. Click into a text field first. A window that does not accept typing has nowhere for the sentence to land.
Per-app modes. You can keep a mode for each app you dictate into. The project has not published a catalog of mode names, so this post will not invent them. The useful idea is that the focused app can carry its own setting instead of making you reconfigure a side panel every time you switch windows.
A personal dictionary. Names, product words, and acronyms go in a dictionary on the Mac so you are not fighting the same misspelling all week.
Install. The command the project publishes is brew install --cask sleekvoice. A build can also be downloaded from GitHub releases (opens in a new tab). If Homebrew cannot find the cask, use the release. The getting started guide walks through macOS Microphone and Accessibility prompts using Apple’s own support articles, because those panes are macOS, not a private SleekVoice screen.
What “free and open source” means here
Free means the price is zero. There is no SleekVoice Pro tier on this site. GPL-3.0 means the license terms are the ones at gnu.org/licenses/gpl-3.0 (opens in a new tab). You can use, study, and share the software. If you distribute a modified version, the license requires you to share the corresponding source under the same license. That sentence is a pointer, not legal advice.
Open source also means the way the project says it is built: in public, with pull requests. The repository is the place to look. On 20 November 2026 that repository did not yet contain a published source tree. Check it when you start, and do not trust a blog post that invents Swift files, Xcode targets, or a contributing guide the repo does not have. Filing an issue and sending a documentation fix are still real contributions. The contribution post is the longer version of that advice.
How it relates to Wispr Flow
Wispr Flow and SleekVoice both try to turn speech into text where you are writing. The differences that are safe to state are the ones either project has written down.
Wispr’s data-controls page (opens in a new tab) says transcription always occurs on the cloud. SleekVoice says Whisper runs locally, including offline. Those sentences are not a benchmark. They are a description of where the audio is processed.
Wispr’s pricing page (opens in a new tab) says Flow is free to start, on Mac, Windows, iPhone, and Android, and that paid plans add team, admin, and support features. SleekVoice is free and macOS-only. A free Wispr plan is not “the same product without the logo.” Cloud transcription remains Wispr’s stated design on the pages reviewed for the comparison.
SleekVoice publishes source. Wispr’s site presents a commercial app. This post does not claim to have audited Wispr’s source practices beyond what those pages show.
What this launch is not claiming
No accuracy percentage. No “hours saved.” No list of companies. No star rating in the page markup. A made-up rating is worse than silence. If you need a number the project has not measured, you will not find it here, and you should be suspicious of a page that offers one without a method.
The text you dictate still enters the app you were using. Local transcription does not stop Slack, Mail, or a synced document from uploading the words after you accept them. That limit is ordinary, and it is worth saying on the day the project launches so the privacy sentence is not stretched past what it covers. The longer version is why local speech-to-text matters.
How to try it
Install Homebrew if the Mac does not have it, then run brew install --cask sleekvoice. If the cask is missing, download a release. Open SleekVoice, allow the microphone when macOS asks, allow Accessibility if macOS asks so the app can type into other programs, click a text field, hold the shortcut the app shows, and say one sentence you can check by eye.
If nothing appears, the docs list the checks in order: cursor, microphone, accessibility, shortcut conflict, then email hello@sleekvoice.app or a GitHub issue with the macOS version.
Questions
Is SleekVoice a Wispr Flow client?
No. It is a separate GPL-3.0 app. It does not sign you into a Wispr account.
Does it work offline?
The project says local Whisper transcription works offline. It does not describe a cloud fallback.
Who makes it?
Bharat Bhavnasi. Email is hello@sleekvoice.app.
- SleekVoice
- voice typing
- macOS
Related reading
- Set up local voice typing on macOS with Whisper
Install SleekVoice, grant the macOS permissions voice typing needs, and use push-to-talk without sending speech to a SleekVoice server.
- Open-source alternatives to Wispr Flow
How to judge an open-source alternative to Wispr Flow, and what SleekVoice, Wispr’s own pages, and Apple Dictation actually say.