What's new in Dictaro

Dictaro now tells you when a new version is out - right in the app, with the release notes one click away.

• You'll hear about every new release - when Dictaro starts, it quietly checks whether a newer version is published and shows a slim bar at the top of the window. No need to keep an eye on the website to know you are up to date.
• See what changed, then decide - the bar links straight to the release notes and to the download, so updating stays your call. Dictaro never downloads or installs anything on its own.
• It never gets in your way - close the bar and it stays closed until the next release. Startup speed and dictation are untouched.

Your API key is now checked the moment you enter it, and richer technical logs make any problem faster to resolve.

• Your key is verified right away - Dictaro checks the key as you save it and shows you which account it belongs to, so a mistyped or expired key is caught immediately instead of surfacing later as something that simply does not work.
• Switch accounts at any time - if you reach your monthly limit, you can now enter a key for another account straight from the dialog and carry on dictating right away.
• More detailed technical logs - the diagnostics export now carries richer technical detail, so should anything ever go wrong, we can help you resolve it faster. Everything you dictate stays on your computer, as always.

Even sharper transcription of long, unbroken dictation - your text finishes cleanly however long you speak without a pause.

• Reliable transcription without pauses - when you dictate a long passage without a single break, Dictaro now processes it in natural passages, so the whole text arrives clean and accurate right to the end - however long you speak without pausing for breath.

Improved recognition and stability of the audio input - Dictaro handles a wider range of microphones more reliably.

• Steadier audio input - Dictaro now copes better with different microphones and audio setups, so recording runs reliably even on less common devices.

A reliability fix: the dictation hotkey no longer stops working while Dictaro is running.

• Your hotkey always stays live - in rare cases the recording hotkey could quietly stop responding while Dictaro kept running, and only came back after a restart. Dictaro now detects this and restores the hotkey on its own within a few seconds, so you never have to restart the app.

Dictaro now learns the words you use most - and dictation feels even more natural, with sharper Slovak and Czech transcription.

• Dictaro learns your vocabulary - when a word you keep dictating isn't recognised, Dictaro quietly notices and, after a few repeats, suggests adding it to your Custom Terms. Your personal vocabulary grows on its own, just from how you speak.
• A calm pause won't cut you off - with Auto-stop enabled, Dictaro now waits a little longer before ending a recording on silence. Take a breath or a short pause mid-sentence and keep going.
• Sharper Slovak & Czech transcription - refined speech detection for Slovak and Czech keeps the engine focused on your voice, for cleaner, more accurate results in both languages.

Your own vocabulary - names, brands, technical terms - now comes through the way you spell it.

• Custom terms, spelled your way - the words you add in Custom Terms now guide the engine as it transcribes, so names, brands and technical terms land correctly instead of being guessed phonetically.
• Matches every form of your words - Dictaro recognises your custom terms even in their inflected forms, with full support for Slovak and Czech word endings, so they stay correct throughout your text.

Dictaro 3.0 - your text appears the moment you stop speaking. Long dictations are now transcribed while you talk, so even multi-minute recordings paste almost instantly.

• Instant results after long dictations - Dictaro transcribes your speech in the background while you are still recording. When you release the hotkey, even a two-minute dictation lands in about a second instead of after a long wait.
• Sharper accuracy on long recordings - speech is processed in natural, pause-bounded passages, which keeps the engine focused and noticeably improves accuracy on multi-minute dictations.
• Same calm flow - nothing changes in how you dictate: press, speak, release. Your text still arrives as one clean block - just much sooner.

Microphone selection fixed for good - picking any microphone in Settings now records from that exact device, even when you plug in or switch devices while Dictaro is running.

• Your chosen microphone, reliably - Dictaro now records through the same proven audio path as the system default, so an explicitly selected microphone no longer produces static, failed dictations or no-speech results on some driver setups.
• Device changes are picked up instantly - Dictaro detects when a microphone is plugged in, unplugged or a dock is connected, and refreshes its audio devices before the next recording. No app restart needed.
• Cleaner saved recordings - locally stored recordings (when enabled) are now saved with the correct audio format header.

Quality-of-life fixes - the floating bar stays visible above other windows, Escape cancels a recording cleanly, and error reporting helps us fix issues faster.

• Floating bar stays on top - the bar surfaces above other windows on every dictation, even after docking, undocking or waking from sleep, and renders at full width right away.
• Escape cancels cleanly - pressing Escape while recording cancels just the dictation, without the key leaking into the app you are working in.
• Faster fixes through error reporting - the app now reports technical errors automatically so recurring issues get diagnosed and fixed sooner. Only technical metadata is sent - never your dictated text.

Steadier bar position across screen changes - the floating bar keeps its place on your current screen when you dock, undock or wake from sleep.

• Steadier bar position across screen changes - the floating bar keeps its place on your current screen when you dock the notebook, undock from a USB-C dock or wake from sleep. No more checking the wrong corner.

Sharper input handling - microphone selection is now honoured precisely across devices, and dictation lands in your active window more reliably.

• Microphone selection, exactly as you pick it - Dictaro now captures from the precise device you choose in Settings across every device type, and gracefully steps aside if a device cannot be opened correctly instead of leaving you guessing.
• Dictation lands where you are - your words paste into the window you were in more reliably, including the very first dictation right after the app starts or your PC wakes from sleep.

Steadier long dictations - recordings of a minute or longer handle edge cases more gracefully when the speech engine produces repetitive output.

• Steadier long dictations - long-form recordings (1+ minute) now handle edge cases more gracefully when the speech engine produces repetitive output. Your dictation lands the way you spoke it.

A meatier release - your microphone selection is now honoured, the silence and hallucination filters cover more languages, and you can browse a per-day plain-text history of every session.

• Microphone that listens - Dictaro records from the exact device you pick in Settings, even when Windows treats a different one as default. Picking a podcast mic or USB headset finally captures from that mic, not the laptop's built-in.
• Smarter silence and hallucination handling - the no-speech filter now recognises common Whisper outros across Slovak, Czech, English, German, Polish and Hungarian, plus catches URL-only fragments, repeated-word loops and stray music notation that occasionally slipped through on quiet captures.
• Browse your dictation history - every successful transcription now lands in a per-day plain-text file at %APPDATA%\Dictaro\sessions\, with a new Session logs shortcut in the SESSION panel that opens the folder directly. Error logs sits next to it for diagnostics.

Reliability polish for dictation paste and localized date display.

• Reliable paste - dictation lands consistently in your focused window across launches and wake-from-sleep, with no first-of-session paste hiccups.
• Localized timestamps - dates and times in user-facing messages now follow the European DD.MM.YYYY HH:MM:SS format.

Account usage updates live while the window is open, with refined behind-the-scenes behavior at startup.

• Live usage refresh - the Account widget keeps the monthly usage count in sync while the window is open, picking up dictations made from other devices on the same account.
• Leaner startup - Dictaro now skips a redundant warm-up step when launching, keeping the first seconds after start lighter and more responsive.

Smoother interaction across the Account panel - refined hover feedback and a more responsive flow when switching between windows.

• Refined hover feedback - the Account panel highlights and shows the pointer cursor consistently while you scan over it, making clickable areas visually unambiguous.
• Responsive window flow - the panel stays interactive after switching focus between Dictaro and other apps, with no need to nudge the window first.

Hover cursor refreshes correctly when returning to the window, and clipboard failures now log to the error log for diagnostics.

• Hover cursor refresh - the clickable Account panel now shows the pointer cursor immediately when you bring the window back to focus, without having to wiggle the mouse first.
• Clipboard diagnostics - if a paste ever fails silently, Dictaro now writes a CLIPBOARD_FAIL or PASTE_FAIL entry to the daily error log so the cause can be tracked.

Fixed clipboard and auto-paste - dictation now reliably lands in the focused window every time.

• Reliable clipboard - switched the clipboard write to Qt's clipboard, which works consistently in the packaged app. Ctrl+V always pastes the last dictation.
• Reliable auto-paste - dictation now reliably types into the focused window (Notepad, VS Code, browser inputs, etc.).

Account panel now refreshes when you open the window and is clickable - one click takes you straight to your portal.

• Auto-refresh on open - every time you bring the window up from the tray, Dictaro pulls fresh tier and usage from the server, so you always see the latest values without restarting.
• Click to open user portal - clicking the Account panel opens dictaro.eu/user in your browser so you can manage keys, plan and account in one move.

Two small fixes on top of v2.5.5 - your plan badge now shows the right name and the Restart App tray action works again.

• Correct plan badge - the Account panel shows the canonical plan name (FREE / INDIVIDUAL / BUSINESS / ENTERPRISE), not the legacy internal label.
• Restart App works - the tray menu's Restart App now reliably relaunches Dictaro instead of doing nothing.

See your account, plan and monthly usage right inside the app - and switch the API key without re-installing.

• Account panel - the main window now shows your name, email, API key name, plan badge (FREE / INDIVIDUAL / BUSINESS / ENTERPRISE) and used / monthly limit, refreshed live as you dictate.
• Change API key - new section in Settings lets you switch keys on the fly when you move between machines or accounts, without uninstalling.
• Quota dialog every time - when your monthly quota is hit, pressing F8 always re-opens the upgrade dialog instead of only showing the floating bar.

Smoother bar transitions - no more flash of the previous result when you start a new recording.

• Clean F8 restart - pressing F8 right after a transcription now jumps straight into Recording without briefly showing the previous "Processed" state.
• Quicker auto-dismiss - the Processed badge fades after 1 second instead of 2, so the bar gets out of your way faster.

Microphone picker stays in sync with Windows Sound Settings - no more app restart to see changes.

• Only enabled microphones - the dropdown now shows exactly what you've enabled in Windows Sound Settings, nothing more.
• Live updates - open the dropdown after enabling or disabling a mic and the list refreshes immediately.
• No loopback noise - Stereo Mix, Line In and similar non-mic capture endpoints are filtered out.

Floating bar visual polish - better proportions, sharper feedback.

• Cleaner state transitions - pill keeps the same height across recording, processing, and result states.
• Refined waveform - bars sit centered with breathing room so the indicator never crowds the edges.
• Sharper close button - repainted X reads clearly against the dark pill.

Clearer feedback after every recording - and a friendly nudge when you reach your monthly limit.

• In-bar status feedback - recording results (Processed, No speech, Failed) now show in the floating bar instead of a system tray balloon.
• Quota-aware bar - when you reach your monthly limit, the bar shows a tap-to-upgrade prompt instead of a confusing error.
• Local error log - failures are recorded to a per-day file you can attach when reporting an issue (Tray menu → Open error log).

Reliability improvements when stopping recordings near the 2-minute limit.

• Stable stop near the 2-minute mark - pressing the hotkey just before the auto-cap no longer triggers a duplicate recording.
• Cleaner no-speech handling - recordings without speech now show a tray notification instead of inserting placeholder text.

The floating recording bar now shows elapsed time and warns you before the 2-minute limit.

• Elapsed timer in the floating bar - see how long you've been dictating at a glance.
• 2-minute soft cap with visual cues - the bar turns amber at 1:40, red for the final 5 seconds, then auto-stops at 2:00.

Microphone selection now lists every active input device on Windows.

• All active mics in the dropdown - devices are deduplicated across MME / DirectSound / WASAPI.
• Cleaner device names - no more "Microphone (...)" wrapper or sample-rate noise.

Self-service API key management lands in the user portal. Re-authentication flow when a key is revoked.

• API key management - generate, copy, and revoke API keys directly in the user portal.
• 401 reauth dialog - when a key is revoked, the app prompts for a new one without restart.
• Copy key ID button next to each active key for support tickets.

UI fully unified to English with an Auto-Enter hint and final polish across menus and dialogs.

• Auto-Enter hint in the floating bar reminds users that pressing Enter sends to the focused app.
• English-only UI for all menus, dialogs, and tooltips (transcription remains 6-language).

Quality-of-life update - taskbar icon and sound toggle.

• Taskbar D icon for quick visual status.
• Sound toggle to mute start/stop chimes during meetings.

Czech and Hungarian diacritic restoration via Hunspell. Better window behavior on Windows.

• Hunspell diacritic restoration for Czech and Hungarian - fills in missing accents Whisper omits.
• Taskbar window - Dictaro now appears in the Windows taskbar like a normal app.
• Mic hot-swap detection - newly plugged microphones appear without restart.

Resizable windows everywhere, with size and position remembered between sessions.

• Resizable main window with persistent size.
• Resizable Custom Terms / Corrections / Filter Words dialogs.
• Recent Clips with full-edge resize and remembered position.
• Auto-close child dialogs when the main window is hidden.