Trace Voice

Help for Windows

How to use Trace Voice on Windows.

Get started

Tested on Windows 11. You can start on the free plan.

  1. Download Trace Voice for Windows, extract the ZIP, open TraceVoice.exe and install.
  2. Sign in with Google, Apple or an email address.
  3. Click where you want to write, press your shortcut and speak.

For install details, see Get started on Windows.

How to dictate

  1. Click where you want to write.
  2. Start speaking with a shortcut.
  3. Release the key or press it again, and the finished text is typed.
  • Drag the voice bar to move it anywhere.
  • Press Esc or the × on the voice bar to cancel. Nothing is typed.
  • The pause button on the voice bar pauses recording; press it again to continue. Audio while paused is not used.
  • Right after input, Undo removes it and Copy copies the result. If the field or text changed, nothing is overwritten.
  • Switch writing style and speech language from the tray menu. To speak in more than one language, add them in Settings (up to three).
  • Trace Voice does not type into password fields.

Shortcuts

These are the default shortcuts.

Right Ctrl
Records while you hold it, and types when you release.
Ctrl + Alt + Space
Press once to start, press again to stop and type.
Ctrl + Alt + P
Polishes the selected text and replaces it.
Ctrl + Alt + R
Shows a translation of the selected text next to the pointer.
Esc
While recording, ends without typing anything. When a translation card is open, closes it.

Change or add keys in Settings → Voice input → Shortcuts. You can also add a shortcut for “Correct last input” there.

Typing into other apps

Works anywhere you can type: browsers, chats, email, notes, terminals and more.

  1. Click a text field in your browser or chat.
  2. Press your shortcut and speak.
  3. What you said is typed into that field.

If Trace Voice could not confirm that the text was typed, a notification appears. Press Copy and paste it into the field.

Real-time input

Real-time input is available on Pro and Max. On Free and Lite, your text appears after you finish speaking.

Words appear faded while you speak and are replaced with polished text when you stop. Translation appears when it is ready.

Turn it off in Settings → Voice input → Real-time input.

In languages with Windows speech recognition, parts heard clearly may appear as a light draft after a short delay. The draft stays on this PC and is never typed, saved or sent. The text that is typed is recognized and polished in the cloud.

Translation

Translation is available on Pro and Max. It covers translating what you say, translating selected text and translating captions.

  1. Right-click the icon in the system tray.
  2. Choose a language under “Translate”.
  3. Speak as usual, and the text is typed in that language.

You can also choose the output language in Settings → Voice input. A translation shortcut with its own target language can be set there too.

Translation covers more than 100 languages in 148 options.

Translate selected text

Select text on a web page or in a message you received and press Ctrl + Alt + R, or choose “Translate selected text” from the tray menu. The translation appears next to the pointer. Use Copy to copy it, or Replace when the text is in a text field.

Change the language by clicking its name on the card, or in Settings → Voice input → “Translate selected text into”. Text already in that language is translated into English. Translations are not saved to History.

Polish text

Select text you have already typed and press Ctrl + Alt + P to replace it with an easier-to-read version.

It also works on text written with another dictation tool or your keyboard.

You can change the key in Settings.

Spoken instructions

Say things like “This is an email” or “Make it a bulleted list” and the text takes that form.

The request itself is not typed.

Input features

On Pro and Max, add templates, snippets and per-app styles in Settings → Voice input → Input features.

Snippets

Say only a snippet’s trigger phrase and the saved text is entered exactly. Nothing is replaced in the middle of a sentence.

Correct last input

Within 3 minutes of an input, use “Correct last input” and say what to change, and that input is rewritten. Add a shortcut yourself in Settings, or use the tray menu. If the target has changed, the result is copied instead of written.

Automatic retry

Voice input that did not finish because the connection dropped is retried automatically and the result is saved to History. It is not typed into any app.

Recordings and meeting notes

Start a recording or import an audio file in Recordings.

  • Recordings are stored encrypted on this PC. By default, only the audio is deleted after 30 days.
  • Transcripts, captions and meeting notes are available on Pro and Max. Recognized characters count toward the monthly allowance.
  • Export notes and transcripts as TXT, PDF, Word or Markdown.
  • Turn on “Also keep dictations in Recordings” to replay and export dictations in Recordings. Playback speed is 0.75 to 2x.
  • Transcription stopped by a lost connection resumes automatically when it returns.

Dictionary

Add names and terms in Dictionary. You can also add readings and aliases.

Search your entries, or add many at once with “Import CSV”.

Sync

Signed in with the same account, your dictionary and settings sync between devices.

Use Cloud sync in Settings → Data → Sync settings to turn sync on or off for this PC. Choose what to sync in Advanced settings.

Syncing text history and recording audio is optional.

Plans and characters

  • Free: 5,000 characters a month
  • Lite: 50,000 characters a month (390 yen a month)
  • Pro: 150,000 characters a month, real-time input, translation, recordings and meeting notes (690 yen a month)
  • Max: unlimited characters, real-time input, translation, recordings and meeting notes (1,580 yen a month)

Prices are for Japan, tax included. They vary by country, and the amount on the purchase screen applies. Yearly billing is also available.

Compare monthly and yearly plans in Account. Use “Manage or cancel plan” to manage your subscription. Eligible first purchases show a seven-day free trial.

Voice input counts the recognized characters toward your monthly allowance. Polishing counts the characters of the original text. Silence and failed parts are not counted.

Unused characters on Lite and Pro carry over to the next period, up to a limit.

One plan works on every device signed in to the same account.

Usage

Open Account → See details to see your usage by day, week, month or year in a chart, split by feature: Voice input, Polishing, Translation and Audio files.

When something goes wrong

Nothing is typed

  • Click where you want to write first, then press your shortcut.
  • If recording does not work, open Windows Settings → Privacy & security → Microphone and turn on microphone access for desktop apps.
  • Trace Voice does not type into password fields.
  • The result is kept in History. You can copy it from there.

The text was copied instead of typed

If Trace Voice could not confirm that the text was typed, a notification appears. Press Copy and paste it into the field. “Correct last input” also copies the result instead of rewriting when the target has changed.

A connection error appeared

On Pro and Max, voice input that did not finish because the connection dropped is retried automatically and the result is saved to History. It is not typed into any app, so copy it from History. A stopped transcription resumes automatically when the connection returns.

The shortcut does not work

  • Check your keys in Settings → Voice input → Shortcuts.
  • If another app uses the same key, choose a different one.
  • Update to the latest version in Settings → General.
  • If it continues, send a report from Account → Report a bug.

If the problem continues, email info@traceforai.com with your app version, your device and what you did.

Apollo Project · Updated November 18, 2026

Translated from the Japanese original (日本語).

← Help Center