Changelog

Setting up VoiceAppear for the first time is now a guided walkthrough in three parts: push-to-talk, hands-free dictation, and a practice email. Each part leads you through the steps one by one.

Why a walkthrough

Many users did not know that hands-free mode exists, because nothing in the app pointed to it. The new setup has you try each mode once, so by the end you have dictated with push-to-talk, used hands-free mode, and written a short sample email with your voice.

Practice out loud

Every step asks you to read a sentence out loud while holding your shortcut. Using the keys once yourself makes them much easier to remember than reading about them in a settings screen.

VoiceAppear now also understands whispering and quiet speech.

The floating bar has a new design:

Kept as minimal as possible

Dictation should fit into what you are doing without distracting from it. The microphone icon is gone and the bar is much narrower: a compact strip that shows the waveform while you talk.

Faster after startup

VoiceAppear now also starts faster after you turn on your PC: it used to take up to a minute, now it is ready within seconds.

You can now turn off the local dictation history. Dictations are then no longer saved on your computer and cannot be restored later. Useful if other people use the same computer.

VoiceAppear now recognizes commands, agents and filenames when you dictate in Claude Code.

Commands come out as commands

Before, saying "slash code review" produced the literal words "slash code review", and command names lost their dashes. The same happened to skills and agent names. You had to correct this manually every time.

Formatted the way Claude Code expects

Now VoiceAppear knows when you are dictating into Claude Code. Commands come out as /code-review, agents and skills keep their exact names, and filenames are spelled the way they should be (camelCase vs. PascalCase, etc.).

Dictate full commands

Since nothing needs correcting anymore, you can dictate a full command, including the slash and its arguments, without typing anything.

VoiceAppear no longer ends the sentence when you pause; punctuation is only added where the sentence actually ends.

You can now set up two different push-to-talk shortcuts and use either one, for example one key at your desk and another on the couch. Every place in the app that shows your shortcut now shows both, and each one keeps its own hands-free combination. Mouse buttons can be part of either shortcut too.

Dictation moved to a new, more accurate speech recognition. Words are transcribed more precisely, and the text appears sooner after you stop speaking.

VoiceAppear now also expects the language your computer is set to, so dictating in your own language works correctly from the first sentence. Other languages are still understood.

A new screenshot button appears next to the floating bar during hands-free dictation. Press it and drag a rectangle over any part of your screen; the screenshot is added to your dictation and arrives in chat apps as a real image.

The Paste button next to the floating bar now handles images as well as text. Copy an image, press Paste while dictating hands-free, and it travels with your text, landing in chat apps as a real image.

VoiceAppear now waits until a wireless headset's microphone is truly listening before it shows as ready, so your first words always make it into the transcript. Headsets that connect through a USB dongle no longer interrupt a dictation.

Everything you say makes it into the transcript, including words spoken after a pause and a short addition right before you let go of the key.

Transcripts now appear within ~300ms on average.

A new model behind it

Dictation moved to a new speech recognition model that is both more accurate and up to three times faster, in every language VoiceAppear supports. The speed does not cost any quality: the new model transcribes more precisely than the old one.

What you will notice

In practice, anything below 350ms feels almost indistinguishable from real-time speed. If you love speed, this is for you.

A short internet dropout while dictating no longer loses anything. The dictation continues on its own once the connection is back, and your text still arrives.

While you are not dictating, VoiceAppear now uses almost none of your computer's resources.

The app window has been redesigned, with a sidebar for Home, Vocabulary, and Style.

The home page now lists everything you have dictated, grouped by day. You can copy any transcript with one click, download it, or delete it, and a retry button runs the original recording through transcription again.

The Vocabulary page holds the words VoiceAppear should always get right (names, jargon, product names). The Style page sets how your dictation is formatted for work, personal, and email contexts, with a live example sentence showing what each option does.

Settings now lets you choose which microphone to dictate with, or leave it on the system default.

Dictation now starts without delay in every app. Pressing your shortcut starts the recording immediately, even in heavy programs and on very large pages.

Transcripts now read naturally where they land. Your words are written down live as you speak, difficult spots around pauses are checked a second time, and punctuation and capitalization are shaped to fit the text your dictation goes into.

While dictating hands-free, a Paste button appears next to the floating bar whenever you have text copied. Press it and the copied text is inserted into your dictation at that point; you keep talking, and it all comes out as one piece of text.

Clicking the floating bar now starts a hands-free dictation, and a stop button on the bar ends it. Right-clicking the bar opens a small menu for opening the app or hiding the bar.

A quick tap of the Windows key opens the Start menu as usual, while holding it to dictate does not. Escape is only intercepted while a dictation is running; the rest of the time it belongs to the app you are in.

Shortcut settings now reject combinations that clash with each other and explain why in plain language.

VoiceAppear's interface is available in English and German. On first start it uses the language your computer is set to, and you can switch in Settings at any time.

Setup now walks you through real dictation practice: a plain text field, a mock email, and a mock chat, each with clear steps to click the field, hold your key, and speak. Quick links let you change your microphone or your shortcut right there.

Dictated text containing quotation marks, tabs, backslashes, or accented letters now pastes exactly as spoken, on both Windows and Mac.

You can now use the middle mouse button or the two side buttons as part of your push-to-talk shortcut. Just click them while recording a new shortcut.

A new Style page lets you choose how VoiceAppear writes in each kind of app: one style for messengers, one for email, one for work tools, and one for everything else. You can turn on all-lowercase, a period at the end, or fewer commas, and a live example shows exactly what your dictation will look like.

VoiceAppear detects what kind of app you are typing into and applies the matching style by itself.

Updates now download in the background and install themselves during a quiet start, so you are never interrupted mid-work. Updating no longer asks for administrator rights.

On Mac you can now hold the Fn key to dictate, on its own or combined with other keys. Holding it for VoiceAppear does not trigger the Mac's own dictation, and the Fn symbol is shown everywhere your shortcut appears.

VoiceAppear is now available on macOS. It installs like any Mac app and works the same way as on Windows: hold your key, speak, and the text is pasted where you are typing.

Setup on Mac walks you through the microphone and accessibility permissions the app needs, and confirms each one as you grant it.

Muting your music while dictating works on Mac too, with wireless headphones and the built-in speakers alike.

When your cursor sits in the middle of a sentence, the first dictated word no longer arrives capitalized as though a new sentence were starting. Names stay capitalized; ordinary words do not.

When your cursor sits right after a word, VoiceAppear adds the missing space itself, so dictated text no longer runs into what is already there.

When you have clearly stopped speaking, the text is finished right away.

Muting while dictating now silences each app individually instead of your whole system, and brings back exactly the apps it muted once you are done. VoiceAppear's own sounds stay audible.

VoiceAppear now cleans up each sentence as soon as you finish saying it, instead of waiting until the end of the dictation. For dictations that need no correction at all, the text appears immediately.

Cleanup also covers German as thoroughly as English, including spoken punctuation and filler words.

VoiceAppear now transcribes continuously while you talk instead of waiting until you let go of the key. The wait after you stop speaking is much shorter, especially for longer dictations.

A new shortcut pastes your most recent transcription wherever your cursor is. Useful when you dictated into the wrong window or need the same text twice.

The wait between pressing your shortcut and being able to speak is now much shorter with AirPods and other Bluetooth microphones.

You can start, stop, and cancel dictations in quick succession, and each one works.

Long dictations now get all the time they need. A two-minute dictation is transcribed in full, just like a short one.

Signing in now happens in your normal web browser. Click sign in, log in there with Google or an email code, and the app picks it up by itself. You stay signed in from then on.

Installing VoiceAppear on Windows is now a single click, with no questions about install locations and no administrator rights needed. The download itself got substantially smaller.

A dictation started right after the previous one now begins immediately.

You can now hold the Windows key, alone or in a combination, to dictate, without the Start menu or Windows search popping open when you let go. Combinations you actually meant for Windows, like the screenshot shortcut, still work as usual.

When you start a sentence over or change your mind mid-sentence, only the version you settled on ends up in the text. Saying "dot" produces a period, and spoken emoji names become the actual emoji.

Transcription got faster: your text arrives sooner after you stop speaking.

Reopening the app takes you straight back in instead of showing the sign-in screen.

While you are dictating, pressing Escape now cancels: nothing is transcribed, nothing is pasted, and your music comes back on.

Transcripts are cleaned up more carefully: filler words and repetitions are removed, question marks are added where they belong, exclamation marks appear only when you say so, and the text no longer ends with a stray period. Emails get the right line breaks for the language you are writing in.

VoiceAppear can now silence your computer's sound the moment you start dictating and bring it back when you are done, so music or a video never bleeds into your recording.

Closing the window no longer quits the app. It keeps running in the background so your dictation shortcut works everywhere, and an icon next to the clock reopens the window or quits the app properly.

VoiceAppear now comes as a signed Windows installer you simply download and run.

Your selected languages and your custom vocabulary are now actually used while transcribing, so names, jargon, and unusual spellings come out the way you set them. Transcripts also arrive properly formatted, with email addresses and file names written correctly.

The entire interface is now available in German: setup, settings, history, vocabulary, and every message in between.

The language list grew from 12 to over 60, including Spanish, Portuguese, Japanese, Korean, Turkish, Polish, Ukrainian, and Vietnamese, plus several German dialects. You can search the list, and your selected languages stay pinned at the top.

Whatever you had copied beforehand (text, formatted text, or an image) is put back on your clipboard right after a dictation, so your clipboard stays untouched.

Holding your shortcut starts the recording and releasing it stops it, whichever keys you use.

Shortcuts can now tell the left and right versions of Ctrl, Alt, Shift, and the Windows key apart, and the shortcut editor shows them by name.

VoiceAppear can now keep listening on its own: tap an extra key while push-to-talk is running and it records until you stop it, useful for anything longer than a sentence. The floating bar shows a stop button while hands-free is active.

Hold a keyboard shortcut, speak, and when you let go your words are pasted as text into whatever app you are in. A small floating bar at the bottom of the screen shows a live waveform while you talk.

The app comes with a history of everything you have dictated, a custom vocabulary for names and jargon, a microphone picker, a choice of 12 languages, and an editor for your keyboard shortcut.