VoiceInk vs Wispr Flow on a Mac: pay once and stay local, or subscribe to the cloud

By Ryleigh Newman · Published October 6, 2026 · Updated October 7, 2026

Pick VoiceInk if you want your voice to stay on your Mac and would rather pay once, and pick Wispr Flow if you want cleanup that follows the app you are in, on a PC and a phone as well, and the cloud does not bother you. VoiceInk is a 7-day trial, then $25 once for one Mac, with its source public under the GPL v3. Wispr Flow always transcribes in the cloud, and its free plan allows 2,000 words a week on desktop before Pro at $15 per user a month, or $12 billed yearly.

Both are serious products. I built YapToText, a free, open-source dictation app that runs on the Mac, so it shows up here as a third option, and you should read that part knowing I made it. I compare what VoiceInk and Wispr Flow publish on their own sites, docs and GitHub, read on October 3 and October 5, 2026. I have not tested either app hands-on against the other, so nothing here ranks them by accuracy.

This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.

On this page: 11 sections

1. Steps
2. Side by side
3. What each one costs
   1. VoiceInk
   2. Wispr Flow
4. On device or in the cloud
   1. VoiceInk
   2. Wispr Flow
5. Cleanup, modes and editing by voice
   1. VoiceInk
   2. Wispr Flow
6. Devices, Intel Macs and languages
7. Which one should you pick
8. Where YapToText fits as a free third option
   1. Step 1: Install it and allow two permissions
   2. Step 2: Tap, talk, tap
   3. Step 3: Turn Wi-Fi off and do it again
9. What each one cannot do
   1. VoiceInk
   2. Wispr Flow
   3. YapToText
10. The short version
11. Questions

Steps

1. Install YapToText from the Mac App Store and open it.
2. In the Get set up card on the Home page, press Continue next to Microphone and Grant next to Accessibility.
3. Click into any text field, tap Right Command, say a few sentences, and tap it again.
4. Check that the words landed at your cursor.
5. Turn Wi-Fi off and repeat the dictation to confirm it works offline.

Side by side

Checked October 3 and 5, 2026, on tryvoiceink.com, VoiceInk's GitHub repository, wisprflow.ai and Wispr Flow's help center: VoiceInk's prices on October 5, Wispr Flow's on October 3. Prices change, so confirm on each vendor's pricing page before you buy.

The row that decides most choices is where the audio goes: VoiceInk keeps it on the Mac unless you connect a cloud provider, and Wispr Flow never does. VoiceInk is a local app you buy once and wire up yourself. Wispr Flow is a cloud service that does the formatting for you and bills by the month. The full comparison of Mac dictation apps adds superwhisper, MacWhisper and Apple Dictation to the same questions.

What each one costs

The two charge in opposite ways. VoiceInk sells a copy of the app, once, by the number of Macs. Wispr Flow rents the service, by the user and the month.

VoiceInk

• The trial. Download from tryvoiceink.com and start the 7-day trial during setup.
• The licenses. Solo $25 for one Mac, Personal $39 for two, Extended $49 for three, and Startup $159 for up to ten. Each is a one-time purchase with lifetime updates and a 14-day money-back guarantee.
• The free route. The source is public under the GPL v3, so you can build it with Xcode. Its README says the compiled, paid app adds automatic updates and priority support.
• Outside the license. A cloud provider you connect for transcription or cleanup bills you on its own terms.

Wispr Flow

• Free. Dictation in any app on every platform, a dictionary and limited Notetaker access. On a Mac or PC the cap is 2,000 words a week, dictation keeps working until 5,000, and past that the next dictation is refused until the reset on Sunday.
• Pro. $15 per user a month, or $12 a month billed yearly, which is $144 a year. It lifts the cap and opens Command Mode.
• Teams. Growth, at $23 per user a month or $18 billed yearly, adds SAML single sign-on, admin control over model training and org-wide HIPAA enforcement with a signed agreement. Enterprise is priced on request.
• Discounts. Wispr's pages describe student offers in more than one way, so check in the app.

My own arithmetic from the prices above, checked October 5, 2026, for VoiceInk and October 3, 2026, for Wispr Flow, in US dollars before any tax. Confirm on each pricing page.

On paper VoiceInk costs a fraction of Wispr Flow Pro, and the gap grows every year. The comparison is not quite fair either way, though: Wispr Flow's free plan may be all a light user needs, its price covers a PC and a phone as well, and VoiceInk's cleanup needs a provider you connect or a Mac with enough memory for Refine. The VoiceInk pricing page works out which license mix costs least, and the Wispr Flow pricing page covers the cap, the student offers and the team plans.

On device or in the cloud

This is where the two differ most, and both say so plainly on their own pages. With VoiceInk the cloud is a door only you can open. With Wispr Flow the cloud is not a setting.

VoiceInk

• By default. Its privacy policy says transcription happens on your device with local models unless you turn on optional cloud services.
• Cloud transcription. Optional, and it sends your audio to the provider you choose, with your own API key.
• AI enhancement. Sends text, not audio, to the provider you connect. Refine, its local cleanup model, keeps it on the Mac.
• Checking the claim. The source is public under the GPL v3, so anyone can read what the app sends.

Wispr Flow

• Always the cloud. Its privacy page says transcription always happens in the cloud, on every plan, and its help center says an internet connection is required. You sign in through a browser during setup.
• Training. Improve the model for everyone is on by default on Free and Pro, and off on Growth and Enterprise. You can switch it off in Settings, Data & Privacy.
• Storage. Dictation Cloud Storage keeps transcripts, audio and history on Wispr's servers until you switch it off, and signing out turns it back on. With both switches off, its help center says no transcription data remains on its servers after processing.
• Context. Context Awareness, on by default, sends the app you are in and the text around your cursor with each dictation, unless a setting its help center calls Privacy Mode is on.
• To be fair to it. Wispr says data is encrypted in transit and at rest, lists SOC 2 Type II and ISO 27001, and keeps zero-data-retention agreements with its third-party AI providers.

Neither company's paperwork is something I can judge for your workplace, and I make no compliance claims for any of the three apps. If a signed agreement is a requirement, Wispr Flow sells one on its Growth plan. VoiceInk's pages do not mention one, and YapToText does not offer one. The Wispr Flow privacy page walks through every switch, and the offline dictation guide shows how to check any app with Wi-Fi off.

Cleanup, modes and editing by voice

Both do more than type what you said. They get there from opposite ends. VoiceInk hands you the parts to wire up yourself. Wispr Flow decides the formatting for you.

VoiceInk

• Speech models. Setup downloads Parakeet V3 by default. Its docs also list Nemotron, Whisper, Cohere, SenseVoice and Apple Speech, which needs macOS 26, plus cloud providers with your own key.
• Cleanup. Through a provider you connect, or VoiceInk Refine, a local model that needs an Apple silicon Mac with at least 16 GB of memory and does one fixed cleanup task.
• Editing a selection. Its docs say Refine cannot use selected text, so rewriting a selection needs Ollama or a cloud provider.
• Modes. They switch by app, website, a spoken trigger phrase or a shortcut. There is also an Assistant Mode for questions, and an output that passes your words to a shell command.
• Your words. Custom vocabulary, word replacements and corrections it learns.
• Files and history. It transcribes audio files, lets you edit a transcript in its history, exports history as CSV, and can delete history on a schedule.

Wispr Flow

• The key. Hold Fn, on Macs with an fn key, and talk. Let go and the words land. One dictation can run up to 20 minutes.
• Cleanup. Backtrack removes filler words, false starts and self-corrections, and automatic formatting turns a spoken first and second into a list.
• Tone per app. Context Awareness and Flow Styles format the message for where it lands, formal in email and casual in chat. Flow Styles work in English only.
• Command Mode. Takes a spoken instruction for selected text on Mac and Windows. It is experimental, off by default, and needs a paid plan or a trial.
• Your words. A dictionary, and it learns names and jargon.
• Files and meetings. Its help center says it takes live speech, not audio or video files. A Notetaker records meetings on Mac and Windows without a bot joining the call, with transcripts stored in Wispr's cloud.

Devices, Intel Macs and languages

Here Wispr Flow reaches further. If you dictate on a PC, an Android phone or an Intel Mac, Wispr Flow is the only one of the two that goes there.

• Other devices. Wispr Flow runs on Windows, iPhone and Android. VoiceInk has an iPhone and iPad app, free during its beta and a one-time purchase after, and its Mac licenses do not cover it.
• Intel Macs. Wispr Flow supports macOS 12 or later on Intel, through its cloud. VoiceInk's pricing page says Apple silicon only, though its install docs suggest cloud transcription on an Intel Mac because its local models do not run reliably there. The guide to dictation on an Intel Mac covers the other choices.
• Installing. VoiceInk comes from tryvoiceink.com, from Homebrew, where the cask is named voiceink, or from the source. Wispr Flow is a download with a sign-in.
• Languages. Wispr Flow lists more than 100. VoiceInk's coverage depends on the model, and its docs say a multilingual model does not cover every language, so check the Language picker before you download one. YapToText lists 29 plus auto-detect.
• Screen readers. Wispr Flow's help center says it is tested with VoiceOver on the Mac. VoiceInk does not say.

Which one should you pick

Start from what you cannot give up. Decide where your voice may go first, then how you want to pay.

• If your audio must not leave the computer, pick VoiceInk with its local models. No setting in Wispr Flow keeps it on the Mac.
• If you want tidy text that fits each app without building anything, pick Wispr Flow. In VoiceInk, cleanup is a provider you connect, or Refine on a Mac with 16 GB.
• If you dislike subscriptions, pick VoiceInk. $25 once covers one Mac, and Wispr Flow bills by the month or the year.
• If you want to read the source, pick VoiceInk. Its code is public under the GPL v3.
• If you dictate lightly and want to pay nothing, Wispr Flow's free plan allows 2,000 words a week on desktop, while VoiceInk's trial ends after 7 days unless you build it yourself.
• If you dictate on Windows, Android or an Intel Mac, pick Wispr Flow.
• If you want to transcribe recordings, pick VoiceInk. Wispr Flow's help center says it does not take audio or video files.
• If your team needs single sign-on, admin controls or a signed agreement, Wispr Flow sells that on Growth or Enterprise. VoiceInk's licenses are by the Mac.
• If you want cleanup, editing by voice and file transcription free, on one Apple silicon Mac, with nothing to connect, neither one gives you all of that for nothing, and that is the gap YapToText fills.

Still torn? VoiceInk's trial and Wispr Flow's free plan cost nothing, so dictate the same paragraphs into each, with a name that usually goes wrong, and count the fixes before the trial runs out. The VoiceInk alternative page and the Wispr Flow alternative page each go deeper on one side.

Where YapToText fits as a free third option

YapToText is the app I built and use all day. It is open source under the GPL-3.0 like VoiceInk, free with no word cap, no account and no analytics, and the speech model and the cleanup model both run on the Mac. It keeps your voice on the Mac like VoiceInk, and does the cleanup without a provider to connect or a plan to pay for.

• Models included. Whisper Large v3 Turbo (Q5), 574 MB, hears you, and Phi-3.5 Mini Instruct, 2.3 GB, cleans up. Both ship in the App Store download.
• Cleanup. Auto mode reads each dictation and picks a treatment: a letter becomes an email, short chat stays as spoken, and everything else gets a careful pass that fixes what was misheard. The Auto mode guide lists the phrases it follows, such as make it formal.
• Editing by voice. Select text in any app, tap Right Option, say the change, and tap again. That is Quick Edit, and it runs on the bundled model.
• Your words. Dictionaries fix a spelling and also prime the speech model, and the Commands page types saved text when you say insert and its phrase.
• Files. Drop an audio or video file on the Utility page and get plain text, on the Mac.

Step 1: Install it and allow two permissions

• You: Get YapToText from the Mac App Store and open it. In the Get set up card on the Home page, press Continue next to Microphone, then Grant next to Accessibility.
• macOS: Asks once about the microphone, then opens System Settings, Privacy & Security, Accessibility, where you switch YapToText on yourself.
• Check: The Get set up card leaves the Home page once nothing is missing.

Step 2: Tap, talk, tap

• You: Click into any text field, tap Right Command, say a few sentences, and tap it again. If VoiceInk or Wispr Flow is listening for the same key, quit it first.
• YapToText: Transcribes on your Mac, shapes the words with Auto mode and types them at your cursor.
• Check: The words are in the field. If they went to the clipboard, Accessibility is not on yet.

Step 3: Turn Wi-Fi off and do it again

• You: Turn Wi-Fi off and repeat Step 2.
• YapToText: Works the same, because both models ship inside the App Store app.
• Check: Run the same test on the other two. VoiceInk's privacy policy says its local models transcribe on your device, and Wispr Flow says transcription needs a connection.

What each one cannot do

Each of the three leaves something out, and not the same thing. VoiceInk asks for money or a build, Wispr Flow leaves out the offline option, and YapToText stays on one Apple silicon Mac with Whisper.

VoiceInk

• No free ready-built app. After 7 days you buy a license or build from source.
• Built for Apple silicon. Its pricing page says Apple silicon only, its install docs point Intel Macs to cloud transcription, and its pages disagree on macOS 14.4 or macOS 15.
• Local cleanup needs memory. Refine wants at least 16 GB and does one fixed task.
• No Windows or Android. Mac, plus an iPhone and iPad app in beta.

Wispr Flow

• No offline dictation. Transcription always happens in the cloud and needs a connection.
• No unlimited free use. 2,000 words a week on desktop before the warning, 5,000 before it stops.
• No editing by voice on the free plan. Command Mode needs a paid plan or a trial.
• No one-time price. Its pricing page lists monthly and yearly billing.
• No file transcription. Its help center says it takes live speech, not files.
• No published source. Not stated.

YapToText

• One kind of computer. An Apple silicon Mac with macOS 14 or later. Nothing for Intel Macs, Windows, iPhone, iPad or Android.
• Whisper only. No Parakeet, Apple Speech or cloud models.
• Fewer languages. 29 plus auto-detect, against more than 100 in Wispr Flow.
• Fewer mode triggers. Modes switch by app, not by website or trigger phrase.
• No meetings or teams. No meeting notes, speaker labels, team admin or signed agreements.
• No hand-edited history. You regenerate an entry instead of editing it.

The short version

• VoiceInk. Open source under the GPL v3, local by default, $25 once for one Mac after a 7-day trial, or free if you build it.
• Wispr Flow. It does cleanup and tone that follow the app, always in the cloud, on Mac, Windows, iPhone and Android. It is free to 2,000 words a week on desktop, then $15 a month or $12 billed yearly.
• The deciding question. Where your voice may go. Check each pricing page before you pay.
• YapToText. Free and open source, with both models inside and cleanup, editing by voice and file transcription included, on one Apple silicon Mac.

Pay once and stay on the Mac, or subscribe and let the cloud do the formatting. If you want neither bill, there is a free third option.

If a price or feature here no longer matches VoiceInk's or Wispr Flow's own pages, or what YapToText does on your Mac, open an issue on GitHub and I will fix it. Every one gets read.

This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.

About the author. Ryleigh Newman is a developer whose hands make typing difficult and who dictates every day with YapToText, the app this site describes. Facts about other apps come from their own sites, with the date they were read. More about the project.

Related

• VoiceInk guides: every page on this site about VoiceInk, by topic
• Wispr Flow guides: every page on this site about Wispr Flow, by topic
• Four Mac dictation apps and a free one, side by side
• VoiceInk pricing: the trial, the one-time licenses, and what is free
• Wispr Flow pricing: what the free plan allows and what Pro adds
• Wispr Flow privacy: where your audio goes, what it keeps, and what you can switch off
• A free VoiceInk alternative, open source with nothing to compile
• A free Wispr Flow alternative that stays on your Mac
• VoiceInk vs superwhisper on a Mac: pay once for open code, or start free
• superwhisper vs Wispr Flow: which one to pick on a Mac

Sources

• https://tryvoiceink.com (checked 2026-10-03)
• https://tryvoiceink.com/pricing (checked 2026-10-05)
• https://tryvoiceink.com/privacy (checked 2026-10-03)
• https://tryvoiceink.com/ios (checked 2026-10-03)
• https://tryvoiceink.com/docs/installation (checked 2026-10-03)
• https://tryvoiceink.com/docs/license-management (checked 2026-10-05)
• https://tryvoiceink.com/docs/introduction (checked 2026-10-03)
• https://tryvoiceink.com/docs/voiceink-refine (checked 2026-10-03)
• https://tryvoiceink.com/docs/recommended-models (checked 2026-10-03)
• https://tryvoiceink.com/docs/local-models (checked 2026-10-05)
• https://tryvoiceink.com/docs/cloud-providers (checked 2026-10-03)
• https://tryvoiceink.com/docs/transcription-history (checked 2026-10-03)
• https://github.com/Beingpax/VoiceInk (checked 2026-10-03)
• https://tryvoiceink.com/docs/mode-triggers (checked 2026-10-06)
• https://tryvoiceink.com/docs/assistant-mode (checked 2026-10-06)
• https://tryvoiceink.com/docs/custom-commands (checked 2026-10-06)
• https://tryvoiceink.com/docs/transcribe-audio-files (checked 2026-10-06)
• https://wisprflow.ai/pricing (checked 2026-10-03)
• https://wisprflow.ai/privacy (checked 2026-10-03)
• https://wisprflow.ai/data-controls (checked 2026-10-03)
• https://wisprflow.ai/students (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/9559327591-flow-plans-and-what-s-included (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/3715544288-weekly-word-limit-not-resetting-troubleshooting-guide (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/2772472373-what-is-flow (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/1036674442-supported-devices-and-system-requirements (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/3152211871-setup-guide (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/4816967992-how-to-use-command-mode (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/4678293671-feature-context-awareness (checked 2026-10-05)
• https://docs.wisprflow.ai/articles/2368263928-how-to-setup-flow-styles (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/5373093536-how-do-i-use-smart-formatting-and-backtrack (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/4841123325-longer-dictation-sessions-now-up-to-20-minutes (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/9609615338-private-cloud-sync-and-data-sharing-preferences-in-wispr-flow (checked 2026-10-05)
• https://docs.wisprflow.ai/articles/3941699399-keyboard-and-screen-reader-accessibility-in-wispr-flow (checked 2026-10-03)
• https://docs.wisprflow.ai/articles/7689167034-can-i-upload-audio-files-to-wispr-flow-for-transcription (checked 2026-10-06)
• https://github.com/ryleighnewman/YapToText (checked 2026-10-03)
• https://apps.apple.com/us/app/yaptotext/id6786382289 (checked 2026-10-03)
• https://wisprflow.ai/whats-new (checked 2026-10-06)

Questions