Guides

Wispr Flow vs Willow Voice on a Mac: two cloud apps at the same price, and a free on-device third option

By ยท Published

Wispr Flow and Willow Voice charge the same for Pro, $15 per user a month or $12 billed yearly, and both turn your speech into text in the cloud, so the choice comes down to the free plan, the extras and what each one keeps. Willow's free Basic plan is unlimited on Frontier Mini, the model Willow itself calls weaker, with 20 Scribe uses a week. Wispr Flow's free plan limits how much you dictate instead: 2,000 words a week on desktop, with a hard stop at 5,000 until the weekly reset.

I built YapToText, a free, open-source dictation app that runs on the Mac, so it appears here as a third option, and you should read that part with my bias in mind. My hands make typing difficult, so I dictate every day. This page compares what each company publishes on its own site and help center, 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.

Download YapToText on the Mac App StoreDownload for free. No account, nothing locked, macOS 14 or later on Apple silicon. Source on GitHub.

On this page: 11 sections
  1. Steps
  2. Side by side
  3. What each one costs
    1. Wispr Flow
    2. Willow Voice
  4. On device or in the cloud
    1. Wispr Flow
    2. Willow Voice
  5. Models, formatting and writing help
    1. Willow Voice
    2. Wispr Flow
  6. Devices, 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: Turn Wi-Fi off and dictate
    3. Step 3: Rewrite what you just said
  9. What each one cannot do
    1. Wispr Flow
    2. Willow Voice
    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. Turn Wi-Fi off, 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 with no connection.
  5. Select what you dictated, tap Right Option, say make it sound more formal, and tap Right Option again.

Side by side

Checked October 3 and 5, 2026, on wisprflow.ai, willowvoice.com and their help centers: Willow's prices on October 5, Wispr Flow's on October 3. Prices change, so confirm on each vendor's pricing page before you buy.

Wispr FlowWillow VoiceYapToText
Free plan2,000 words a week on desktop and 1,000 on mobile; on desktop, dictation is refused past 5,000 until the weekly resetBasic: unlimited use of Frontier Mini, 20 Scribe uses a week, 5 minutes per dictationEverything, with no word cap
Pro$15 per user a month, or $12 billed yearly$15 per user a month, or $12 billed yearlyNone; tips are optional and buy no features
Team plansGrowth $23, or $18 billed yearly; Enterprise on requestBusiness $35, or $28 billed yearly; Enterprise customNone
What Pro unlocksNo weekly cap, and Command ModeFrontier Pro, unlimited Scribe, 8-minute dictations, offline dictation and a memory of your writing styleNothing is locked
Where speech becomes textWispr's cloud, always, per its privacy pageCloud-based AI models; offline dictation listed under ProOn the Mac
Kept by defaultDictation Cloud Storage on until you switch it off; model training on by default on Free and ProPrivate Mode on: no audio stored on Willow's servers, transcript history on your deviceNothing leaves the Mac
Longest single dictation20 minutes5 minutes on Basic, 8 on paid plansA 30-minute safety stop, which still saves the dictation
Default keyHold Fn, on Macs with an fn keyHold Fn, or double-tap it to keep listeningRight Command, tapped to start and stop
Writing or editing by voiceCommand Mode, experimental, for paid plans and trialsScribe writes a message from a description or rewrites a selection; 20 uses a week on BasicQuick Edit, free
DevicesMac, Windows, iPhone, AndroidMac, Windows, iPhone; Android coming soonMac only
Mac requirementmacOS 12 or later, Apple silicon or IntelNot statedApple silicon, macOS 14 or later
LanguagesMore than 100More than 10029, plus auto-detect
Open sourceNot statedNot statedYes, GPL-3.0

The Pro prices match to the cent, so the real decision is the free plan and what each company does with your audio. Wispr Flow rations how many words you dictate for free; Willow rations which model hears you. The full comparison of Mac dictation apps adds superwhisper, MacWhisper and VoiceInk, and the roundup of Wispr Flow alternatives ranks Willow next to the other cloud apps.

What each one costs

For one person, both companies offer the same two choices, free or Pro at the same price. The free plans are where they split. Wispr Flow's free plan is limited by a word count; Willow's is unlimited on the weaker model.

Wispr Flow

  • Free. Dictation in any app on every platform and a dictionary, plus limited Notetaker access. On a Mac or PC, 2,000 words a week is the warning; dictation keeps working to 5,000, and past that the next dictation is refused until the reset on Sunday. Phones share a separate pool of 1,000 words.
  • Pro. $15 per user a month, or $12 a month billed yearly. It lifts the cap and opens Command Mode. The pricing page says you need no credit card to start and that Pro can be canceled at any time.
  • 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.

Willow Voice

  • Basic. Free, with no credit card: unlimited use of Frontier Mini, limited personalization, 20 Scribe uses a week and dictations of up to 5 minutes.
  • Pro. $15 per user a month, or $12 billed yearly, for Frontier Pro, which Willow calls its smartest model, with faster priority transcription, unlimited Scribe, longer dictations, a memory of your writing style and offline dictation.
  • Business. $35 per user a month, or $28 billed yearly, adds enforced privacy mode with zero data retention, SOC 2 Type II, enforced HIPAA compliance, and admin and team controls. Enterprise is priced through Willow's sales team.
  • Discounts. Willow's help center says students get 30 percent off Pro by emailing its support team with proof of enrollment.

One flag on Willow. On October 5, 2026, a plan chart in Willow's help center did not match its pricing page: the chart showed a free plan with 2,000 words a week, and the pricing page showed no word cap on Basic. This page follows the pricing page, but watch for a limit in the app before you count on unlimited.

My own arithmetic from the prices above, checked October 5, 2026, for Willow and October 3, 2026, for Wispr Flow, per user and before any tax. Confirm on each pricing page.

One year ofWispr FlowWillow Voice
Pro, paid monthly$180$180
Pro, billed yearly$144$144
Team plan, paid monthly$276, Growth$420, Business
Team plan, billed yearly$216, Growth$336, Business

For one person the bill is identical. For a team, Growth and Business list different features, so compare those, not only the totals. The Wispr Flow pricing page and the Willow Voice pricing page break down every plan.

On device or in the cloud

Both are cloud apps: each company says its transcription runs on cloud models. Where they differ is what is kept afterwards, and what you have to switch off yourself. Willow starts private and lets you opt in to sharing; Wispr Flow starts by keeping more, and gives you the switches.

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.
  • 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. Wispr says it skips standard macOS password fields and banking apps, though custom or web-based password fields may be read like normal text fields.
  • Safeguards. 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.

Willow Voice

  • Cloud models. Willow's help center says it uses cloud-based AI models, on cloud infrastructure in the United States.
  • Private Mode. On by default and available on every plan. The help center says no dictated text is collected, transcripts and context are saved only on your device, and audio is kept on the device only so it can be transcribed again.
  • Sharing. If you switch to Help Willow Improve, Willow collects your transcript text with your account anonymized.
  • Offline. The help center lists offline dictation as part of Pro. Willow does not say which model runs offline, so ask Willow if that matters to you.
App Store poster headed Nothing leaves your Mac, over YapToText's AI Models page with both models marked Included.
Figure 1. The third answer to where your voice goes: in YapToText both models ship inside the App Store app and show as Included, so there is no cloud default to change.

I make no compliance claims for any of the three apps, and only your organization can judge the paperwork. If you need a signed agreement, Wispr Flow offers one on Growth and Enterprise and to individuals in the app, Willow Business lists enforced HIPAA compliance, and YapToText offers none. The Wispr Flow privacy page walks through each switch, and the offline dictation guide shows how to test any app with Wi-Fi off.

Models, formatting and writing help

Both apps do more than type what you said, and here they really are different products. Willow will write the message for you; Wispr Flow tidies the one you said.

Willow Voice

  • Two models. Frontier Mini on Basic, and Frontier Pro on paid plans. Willow's own pricing page is the source for calling Mini the weaker one.
  • Formatting. Willow is context-aware, and in an email app it formats the greeting, the body and the closing.
  • Scribe. Describe what you want to say and Scribe writes the message in your style, including replies that read what is on your screen. It has its own shortcut, Fn plus Control, and it also rewrites text you select.
  • Your words. A personal dictionary on every plan, and on Pro a memory of your writing style.

Wispr Flow

  • The model. Not stated by Wispr.
  • 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 shape 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.
  • Meetings. A Notetaker records meetings on Mac and Windows without a bot joining the call, with transcripts stored in Wispr's cloud. Willow does not list meeting notes.

If starting a message is the hard part, Scribe has no match here. If you want your own words kept and only cleaned, Backtrack is closer, and so is YapToText's Auto mode.

Devices, Macs and languages

Both follow you off the Mac, which YapToText does not. If you dictate on an Android phone, Wispr Flow is the one of the two that is there today.

  • Other devices. Wispr Flow runs on Windows, iPhone and Android. Willow runs on Windows and iPhone, with Android listed as coming soon on its download page.
  • One subscription. Willow's help center says one Pro subscription covers Mac, Windows and iPhone, as long as you sign in the same way, Google, Apple or email, on each.
  • Intel and older Macs. Wispr Flow supports macOS 12 or later on Apple silicon or Intel. Willow states no Mac requirement, so check before you rely on an older Mac. The guide to dictation on an Intel Mac covers the choices there.
  • Languages. Both list more than 100. Wispr's Flow Styles are English only. YapToText lists 29 plus auto-detect, and the languages guide names them.
  • Screen readers. Wispr Flow's help center says it is tested with VoiceOver on the Mac. Willow does not say.

Which one should you pick

Start with the thing you cannot give up, then let the rest decide. If you will pay, the price is a tie; if you will not, the free plans point in opposite directions.

  • If you want to dictate a lot without paying, pick Willow Basic. Its pricing page shows no word cap, while Wispr Flow's free plan stops at 5,000 words a week on desktop. Watch for the 2,000-word chart in Willow's help center.
  • If you dictate a little and want Wispr Flow's cleanup without paying, pick Wispr Flow Free. What limits it is the weekly word count, plus Command Mode staying on paid plans.
  • If you want the app to draft messages from a rough description, pick Willow and its Scribe.
  • If you talk in long stretches, pick Wispr Flow. One dictation can run 20 minutes, against Willow's 8 on paid plans.
  • If you use Android, pick Wispr Flow. Willow lists it as coming soon.
  • If you want little kept on servers without touching settings, Willow's Private Mode is on from the start. In Wispr Flow you switch off training and cloud storage yourself.
  • If your audio must not leave the Mac, neither one's default does that, and that is the gap YapToText fills.

Still torn? Both have a free way in, so dictate the same three paragraphs into each, with a name that usually goes wrong, and count the fixes. The Willow Voice alternative page covers Willow in more depth, and superwhisper vs Wispr Flow sets Wispr Flow against a local app.

Where YapToText fits as a free third option

YapToText is the app I built and use all day. It is free with no paid tier, no account, no analytics and no word cap, the code is public under the GPL-3.0, and both of its models run on the Mac: Whisper Large v3 Turbo (Q5) hears you, and Phi-3.5 Mini Instruct cleans up. There is no weaker free model and no count to watch, because there is only one tier.

  • 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 and rewrites nothing.
  • Editing by voice. Quick Edit rewrites a selection in any app, with no weekly count. It does not write a message from a description the way Scribe does.
  • Your words. Dictionaries fix a spelling and also prime the speech model, so a name is heard right in the first place.

You can try it beside either app in a few minutes. Quit the other one first so two apps are not listening for a key at once.

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: Turn Wi-Fi off and dictate

  • You: Turn Wi-Fi off, click into any text field, tap Right Command, say a few sentences, and tap it again.
  • YapToText: Transcribes on your Mac, shapes the words with Auto mode and types them at your cursor.
  • Check: The words land with no connection. If they went to the clipboard, Accessibility is not on yet.

Step 3: Rewrite what you just said

Say: Select the sentences you dictated, tap Right Option, say this, and tap Right Option again.

make it sound more formal

What happens: The Quick Edit card listens, applies the edit on your Mac and says Done, and your selection is replaced.

App Store poster headed Edit anything by voice: the Quick Edit card listening, applying the edit, and done.
Figure 2. Step 3 on screen: Quick Edit listens for the change, applies it on the Mac and says Done, with no weekly count and no paid plan.

What each one cannot do

All three have gaps, and they point in different directions. Wispr Flow counts your free words, Willow holds back its better model, and YapToText stays on one Apple silicon Mac.

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.

Willow Voice

  • The weaker model on Basic. Frontier Pro is for paid plans.
  • Rationed Scribe and short dictations on Basic. 20 Scribe uses a week and 5 minutes per dictation.
  • No offline dictation on Basic. The help center lists it under Pro.
  • No Android yet. Listed as coming soon.

YapToText

  • One kind of computer. An Apple silicon Mac with macOS 14 or later; nothing for Intel Macs, Windows, iPhone or Android.
  • Fewer languages. 29 plus auto-detect, against more than 100 in both of the others.
  • No message drafting. It cleans up what you said and edits a selection, but it does not write a message from a description.
  • No meetings or teams. No meeting notes, team admin or signed agreements.
  • No screen reader test pass yet. It announces every state to VoiceOver, but I have not run a full VoiceOver test pass.

The short version

  • Wispr Flow. Free to 2,000 words a week on desktop, then $15 a month or $12 billed yearly; always in the cloud, with Android and 20-minute dictations.
  • Willow Voice. Unlimited on the weaker model for free, Frontier Pro at the same $15 or $12; Private Mode on by default, and Scribe to draft for you.
  • Both. Mac, Windows and iPhone, more than 100 languages, and the cloud. Check each pricing page before you pay.
  • YapToText. Free and open source, with cleanup and editing by voice on one Apple silicon Mac, and your voice kept there.

Same price, different free plans; if you want neither the cloud nor the bill, there is a third option.

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

Download YapToText on the Mac App StoreDownload for free. No account, nothing locked, macOS 14 or later on Apple silicon. 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.

Sources

Questions

Is Wispr Flow better than Willow Voice?

It depends on what you need. Wispr Flow runs on Android, allows 20-minute dictations and has meeting notes, while Willow's free plan is unlimited on its weaker model, its Private Mode is on by default, and Scribe drafts messages from a description.

Which is cheaper, Wispr Flow or Willow Voice?

For one person they cost the same: Pro is $15 per user a month or $12 billed yearly on both, checked October 3 and 5, 2026. For teams, Wispr Flow Growth is $23 or $18 billed yearly, and Willow Business is $35 or $28.

Is Willow Voice free with no word limit?

Willow's pricing page lists Basic as unlimited on Frontier Mini, the model Willow calls weaker, with 20 Scribe uses a week. A plan chart in its help center showed 2,000 words a week on October 5, 2026, so watch for a limit in the app.

Do Wispr Flow and Willow Voice work offline?

Wispr Flow does not: its privacy page says transcription always happens in the cloud. Willow's help center says it uses cloud-based models and lists offline dictation as part of Pro. YapToText works offline on every dictation.

Do Wispr Flow and Willow Voice work on Android?

Wispr Flow does. Willow runs on Mac, Windows and iPhone and lists Android as coming soon. YapToText is a Mac app only.

Is there a free alternative to Wispr Flow and Willow Voice?

Yes. YapToText is free and open source, runs its speech and cleanup models on your Mac, and edits selected text by voice. It needs an Apple silicon Mac with macOS 14 or later.