Updated September 14, 2026 · By Sumbat.T

Spokenly for Windows: Does It Actually Work on a PC?

Dictation built for Windows, not ported to it

Talk into any text field on your PC

BlabbyAI is a Windows desktop app and a Chrome extension. Press Ctrl+Space, speak, and the text lands where your cursor is. History keeps every recording on your own disk. 60 credits a week free, no card.

Add BlabbyAI to Chrome

The short version, by what you actually need

"Does it work on Windows" is rarely the real question. The real question is which tool you should install this afternoon, and that depends on whether you care most about running offline, about having a mature client, or about getting past a locked-down work machine. These six lines are the whole page compressed:

Why this question exists at all

Most software does not generate a "does it run on Windows" search. Spokenly does, and for a specific reason: the app was a Mac and iPhone product first, it got its press coverage in that period, and the coverage never caught up.

Look at what a search for the product name returns today. The App Store listing. A thread in r/macapps. MacUpdate, which serves Mac software. A Lifehacker piece headlined as a free Mac app. A blog post describing it as a macOS and iOS app. Spokenly's own subreddit, whose description reads AI-powered dictation for Mac & iOS. A reader scanning that page would conclude the app is Apple-only, and they would be wrong.

The sharpest example is a review that currently ranks on the second page for this exact query. Dated 26 May 2026, it tells readers in as many words that there is no Windows download link on the vendor's site and that the pricing and feature grid only cover Mac and iPhone. That was already wrong when it was published: Spokenly's first Windows release, 0.1.0, shipped on 3 May 2026, twenty-three days earlier.

What actually ships for Windows today

Here is the vendor's own Windows page, which is the primary source for everything in this section. It states the version, the date and the system requirements on the same screen as the download button.

Source: Spokenly's Windows page, checked 14 September 2026

The specifics, all read from the vendor's own pages on 14 September 2026:

• Version 0.3.20, updated August 2026.
• Windows 10 and 11, 64-bit. There is no 32-bit build.
• 29 MB installer, direct download from the vendor's site. It is not distributed through the Microsoft Store.
• Local models: Whisper and Parakeet, running offline on your own machine.
• Cloud models: Deepgram, OpenAI, Soniox and ElevenLabs, with 11 providers in total and a bring-your-own-key option.
• AI text processing via GPT-4, Claude, Gemini, Grok and local models through Ollama.
• A global hotkey that inserts text at the cursor in any Windows application. Google's AI Overview for this query gives the default as Ctrl+Space, though the vendor's Windows page does not state it, so treat that one as second-hand and check it in settings after installing.

That is a complete dictation product, not a stub. Whatever else this page says about maturity, nobody should be left with the impression that the Windows build is a placeholder or a waitlist.

The number that nobody else prints: 0.3.20 against 2.29.4

Spokenly publishes separate release pages for each platform, and reading them side by side gives the clearest possible answer to how the Windows build compares. This is the part of the story that is missing everywhere else, and it takes about a minute to verify. The last row is the contrast that matters for a Windows reader.

Two readings of this are wrong and one is right. The wrong ones are "Spokenly does not support Windows", which the download page disproves, and "Spokenly supports Windows and macOS equally", which the version numbers disprove. The right reading is that Windows is a real, actively developed, four-month-old client that the developer has not yet called 1.0.

You do not have to take that inference from us, because the vendor states it directly on the Windows page:

That is a straightforward thing for a vendor to publish about its own product, and it deserves credit. It is also exactly what a Windows buyer needs to know before spending an afternoon on a migration.

The release cadence is the good news

Twenty-four Windows releases between 3 May and 29 August 2026 is roughly one every five days. The release notes read like genuine platform work rather than version bumps: one says "Dictation now inserts text more reliably, including in terminals", another that "Parakeet models now use your GPU and are much faster". Those are the two things that most often go wrong with a young dictation client on Windows, and both have been worked on.

So the honest forecast is that the gap narrows. If you are reading this months after publication, check the version number again, because on this cadence it will have moved.

What it costs on Windows, and the number the pricing page hides

Spokenly's pricing is unusually generous at the free end and slightly misleading at the paid end, in a way that is worth a paragraph.

Source: Spokenly's pricing page, checked 14 September 2026. Note the Yearly toggle above the cards.

The free end is real. Spokenly lists Local Models at "$0 forever" with "Unlimited use", "Works completely offline" and "No usage limits", requiring no account. A second free column lets you bring your own API keys, where "Spokenly charges nothing" and you pay OpenAI, Deepgram or Groq directly. Its own comparison note puts it well: "Everything local is free forever in every plan. Pro only adds the cloud."

The paid end needs care. The card reads $8.33/month, but the toggle above it is set to Yearly, and the small print says billed $99.99 per year. Pay month to month through the Apple App Store and the listed in-app purchase is $9.99, with a six-month plan at $49.99. So depending on where and how you buy, the same Pro tier is $8.33 or $9.99 a month, a 20 percent spread that no single page states in one place.

Spokenly against the other Windows options

Every price below was read from the vendor's own pricing page on 14 September 2026, and monthly and annual rates are stated separately wherever they differ, because in this category they usually do.

Tool	Windows	Free tier	Paid
Spokenly	Yes, 0.3.20, direct download	Unlimited offline local models, no account	$8.33/mo billed yearly, $9.99 monthly on the App Store
BlabbyAI	Yes, Windows app plus Chrome extension	60 credits a week, no card	$8.49/mo
Wispr Flow	Yes	2,000 words a week on desktop	$15/mo, $12/mo billed annually
Superwhisper	Yes, but the fully local setup is macOS only	3,000 Pro words, then free tier stays	$8.49/mo

The thing to notice is that the prices barely differ. Spokenly Pro, BlabbyAI and Superwhisper all sit within sixteen cents of each other per month. Wispr Flow is the outlier at nearly double. So on Windows this is not a price decision, and anyone framing it as one is padding a comparison. It is a decision about whether you want offline models you manage yourself, or a product that works the moment it opens.

BlabbyAI: Windows as the product, not the port

The version numbers above frame the real choice on this platform. You can run a four-month-old Windows client, or you can run one where Windows was the target from the start.

BlabbyAI is built for Windows rather than brought to it. The toolbar sits above the taskbar, the global shortcut is Ctrl+Space and works from anywhere in Windows regardless of which application has focus, auto-launch is on by default so the shortcut is live after every boot, and onboarding is two steps with no account and no sign-in. There are no models to choose and nothing to download before it works.

Source: BlabbyAI shortcut settings on Windows

History keeps the audio on your disk

This is the feature that answers the same worry Spokenly's local models answer, from a different direction. When a recording stops, BlabbyAI writes the audio to your own disk before transcription begins, along with the resulting text. If the network drops, the server fails, or you cancel halfway, the audio is still there and you can re-run it later, optionally through a different mode or language. The default cap is 500 entries and it is configurable. None of that data touches a server.

Custom modes are the part with no Spokenly equivalent

Spokenly has AI text processing, and it is a real feature: connect a model and have your transcript cleaned up. BlabbyAI's custom modes work differently in one respect that matters, which is that the instruction is free-form and yours. You write it, so it can do approximately anything a language model can do to a block of text.

A worked example. You create a mode whose instruction is: "Translate what I said into English, fix any grammar errors, and format it as a short email. Whenever I say 'new paragraph', insert an actual paragraph break instead of writing those words." Assign it a shortcut. Now you speak in your own language, at your own pace, saying "new paragraph" where you want one, and a clean English email lands in the text field. That is four separate operations in one keypress.

Source: BlabbyAI mode editor, where the instruction is free text

That mechanism also covers the classic dictation-command use case without a built-in command list. Because the model receives the full literal transcript, including words you spoke as commands, you can define your own vocabulary: "when I say bullet point, start a new bulleted list item", "if I say scratch that, delete the sentence I just said". You define the vocabulary rather than learning someone else's.

The Chrome extension, for machines you do not control

There is a case this whole comparison misses if we only talk about desktop installers. Spokenly's Windows build is a direct download from the open web and is not on the Microsoft Store. On a managed work laptop, that is frequently the end of the conversation, whatever the software's merits.

The BlabbyAI Chrome extension installs from the Chrome Web Store with no desktop installer at all. It shows a small bubble next to whatever text field you have focused, and it works in Gmail, Google Docs, Notion, Jira, Slack in the browser, a CRM, a ticketing system, anything with a text box. It needs no administrator and no executable. For a lot of readers on work machines, that is the difference between using a dictation tool and not using one.

"Spokenly alternative for Windows" is the question behind the question

Something revealing shows up on the search results page for this topic. Google's top related search, repeated in every block on the page, is not "Spokenly download" or "Spokenly pricing". It is "Spokenly alternative for Windows". And the second organic result is a Reddit thread titled exactly that, with people answering it.

That tells you what has been happening. Windows users hear about Spokenly, land on coverage written for a Mac audience, conclude it is not for them, and go looking for something else. Some of them are leaving a product that would in fact have run on their PC. Others have tried the 0.3.20 build and want something further along.

If you are in the first group, the download is on the vendor's Windows page and the local models cost nothing, so the question is answered. If you are in the second group, the Windows-first option is BlabbyAI at $8.49 a month, where the Windows app and the Chrome extension are the products rather than ports, with Wispr Flow the remaining contender at roughly double that.

How to decide in about two minutes

1. Does the audio have to stay on your machine? That is the one requirement a local model answers and a hosted one does not, and Spokenly's free tier is built for it, provided your PC has a GPU worth the name.
2. Can you install executables on this machine? If not, the desktop question is settled for you and a browser extension is the route.
3. Do you want to manage models, or do you want it to work when it opens? Local models mean downloads, hardware dependence and a settings screen. Hosted means none of that and a monthly bill.
4. Do you need the output reshaped, not just transcribed? If you regularly dictate in one language and need to send another, or you want a fixed format every time, a free-form custom mode does that in the same keypress.
5. How much does a 0.x version string bother you? This is a real question and only you can answer it. The build is active and the vendor is candid about its maturity.

Frequently asked questions

Sources

• Spokenly Windows page, download page, pricing page and per-platform release pages, all read 14 September 2026.
• Apple App Store listing for Spokenly, in-app purchase prices and version 1.12.3, read 14 September 2026.
• Wispr Flow pricing page, read 14 September 2026.
• Google search results and AI Overview for "spokenly" and "spokenly windows", United States, 14 September 2026.