Dictpad guide — one record button turns speech into editable notes that stay on this device

When talking is faster than typing

The things you want to write down right after a meeting, the sentence that arrives while you are walking, tomorrow’s to-dos you do not want to lose before sleep. All of these are far faster to say than to type. Yet when you actually try to “just say it”, the road is oddly narrow. The mic key on a phone keyboard only types into whichever box happens to be open, and when you finish talking the words are scattered across some app’s form field. It was a keyboard, not a notepad.

So people go looking for a “dictation app”. The app stores have plenty. Open one, though, and the first screen is often a checkout: a few dollars a week, a three-day trial, then auto-renew. The core feature, speech recognition, already ships inside the browser and the operating system; what you are paying for weekly is the notepad on top. Most of them also want an account and want your recordings on their server.

The web has had one good answer for years. A browser dictation notepad like Speechnotes lets you press the mic and watch text pile up on the page, say “period” to get a full stop, and copy it out when you are done. No account, no checkout. But that notepad has no Korean interface. You can pick English or Japanese, yet there is no Korean UI, and the punctuation commands do not understand Korean either: say “마침표” and you get the word, not the mark.

What people want is simple. Press one button, talk, watch the words land in a notepad you can fix right away, keep that text on your own device, and never see a screen asking for money every week. And have “period” and “new line” work in Korean and Japanese too, not only in English.

What Dictpad does

Dictpad is that notepad. Open it and you see a big coral Record button and a ruled notepad. Tap Record and speak: words that are not final yet show first in the sky-blue strip under the button, and the moment a sentence is final it is appended to the end of the notepad. The notepad is an ordinary text box, so while recording you can move the cursor, fix a word, delete a line or type. Tap Stop and the microphone is off; tap Record again and a fresh session starts cleanly.

Punctuation is spoken. With Korean recognition, “마침표”, “쉼표”, “줄바꿈”, “단락” and “물음표” work; with English, “period”, “comma”, “new line”, “new paragraph” and “question mark”; with Japanese, “句点”, “読点”, “改行”, “段落”. Say a command in the middle of a sentence and the mark lands right there. Japanese gets full-width “。” and “、”.

You can keep several notes. Make one with “New note”, give it a title, delete it when it is done. Picking a note loads its text into the notepad, and whatever you type or dictate is saved on this device within a fraction of a second. When you are finished, “Copy” puts the text on the clipboard to paste anywhere, “Download .txt” saves a file, and “Export JSON” packs every note into one backup file.

Banner, mic deck, notepad, voice commands, notes to self, backup

• Amber banner at the top: the promise that notes stay on this device, with no login, no ads, no paid subscription, and that this is an in-browser notepad rather than a system keyboard. Change the screen language (ko/en/ja/de) at the top right.
• Mic deck: a status line (Idle, Starting, Listening) with the recognition code, the “Recognition language” picker (ko-KR, en-US, en-GB, ja-JP and close to thirty in all), the big “Record / Stop” button, and under it the sky-blue strip where not-yet-final speech appears. If the microphone is denied or the connection drops, an explanation appears right there.
• Notepad card: the note picker, “New note”, “Rename”, “Delete”. The big ruled text box, an “Autosaved on this device” mark and a character count. Below it “Copy”, “Download .txt”, “Export JSON”, “Import JSON”, “Delete everything”.
• Voice punctuation card: the command words for the recognition language you picked, next to the mark each one inserts (., , ↵, ¶, ?). Change the language and the card changes with it.
• Good-to-know card: that Chrome on Android starts recording with one tap and keeps the screen awake while recording, that recognition uses the browser’s speech service and so usually needs a connection, that this is not a system keyboard typing into other apps, and that you can edit the text while recording.
• Below that the row of promise chips and the footer (Privacy, Terms, Guide, try-dabble.com). The “Guide & Feedback” button at the bottom right opens this guide page in a new tab.

Where your data stays

Notes, their titles, which note is open and the recognition-language setting all live inside this browser on this device. Nothing goes up to a Dictpad server and there is no account. That is also why opening it in a different browser on the same device shows an empty notepad.

One thing should be said plainly. Turning speech into text is done by the browser’s own speech recognition. Chrome and most other browsers send the sound while you are recording to the browser maker’s speech service and get text back, and that part follows that browser’s privacy policy. Dictpad never receives that audio and never stores it. When you are not recording, the microphone is off. This is also why recognition usually needs an internet connection, while the notepad itself opens offline from the second visit on.

Before you switch devices or clear browser data, press “Export JSON” and keep the file. “Import JSON” on the new device brings every note back with its title. The file is created on this device and sent nowhere.