Saypen privacy policy

Effective date: 29 October 2026 · Applies to: Saypen – Private Voice Typing, version 1.0.0 and later, an extension for Google Chrome on desktop computers, published by Iliria Studio.

Summary

Saypen types what you say into the text field you are using. Your speech is turned into text by Chrome’s built-in speech recognition, on your computer. Saypen itself makes no network requests: it has no servers, no accounts, no analytics and no ads. It never records, saves or sends your audio, and it keeps no history of what you dictate. Your settings are stored only on this computer.

One download is involved, and Chrome makes it, not Saypen: the first time you use a language, Chrome downloads that language’s speech model from Google, after you click Download on Saypen’s setup page (see §2).

No network requests
No account
No analytics
No ads
No audio stored
No history

1.What Saypen handles, and why

Saypen handles data only while you use it, and only for its single purpose: typing what you say into the text field you chose.

Data	When	Why, and what happens to it
Microphone audio (your voice)	Only while voice typing is on, or while the “Try it here” box on Saypen’s setup page is listening	It goes straight to Chrome’s on-device speech recognizer, which turns it into text on your computer. Saypen never records, saves or sends audio.
The recognized text	While voice typing is on	It is typed into the text field you chose, on the page you are using. While you speak, Saypen’s bar on that page shows the words as they are recognized.
The few characters just before the cursor in that field	Each time a phrase is typed	To decide whether to add a space or a capital letter. They are read at that moment and not stored.
Which element on the page has focus	When voice typing starts and while it runs	To find the text field you are using, including inside frames and shadow DOM.
The state of the current voice-typing session, including the pieces of text typed in it	Only while voice typing is on	Kept in Chrome’s session memory (`chrome.storage.session`) so that the voice command “delete that” can remove the last phrase, even if Chrome restarts Saypen’s background script. It is memory only and never written to disk. It is removed as soon as voice typing stops, and when the browser starts.
Text that could not be typed	When no text field was selected	Shown in Saypen’s bar on that page with Copy and Close buttons, so it is never lost silently. It stays there until you copy or close it, and is not kept anywhere else.
Your settings: language, automatic punctuation and capital letters on or off, voice commands on or off, silence timeout, start and stop sounds on or off	When you change them	Saved in `chrome.storage.local`, on this computer only and never synced, to remember your choices.

Saypen does not read other content of the page, cookies, form data or your browsing history, and it stores nothing from the pages you use.

2.What Chrome does

Speech recognition. Saypen uses Chrome’s built-in speech recognition (the Web Speech API) with on-device processing switched on. Chrome turns your audio into text on your computer.
The language download. The first time you use a language, Chrome needs that language’s speech model, which takes about 200 MB on disk. You start the download by clicking Download on Saypen’s setup page, and Chrome then downloads the model from Google. Saypen does not download it; it only shows you whether it is ready. The model is shared by all websites, and Chrome manages it and keeps it for later use. After that, recognition works without an internet connection.
Microphone permission. Chrome asks for your permission before Saypen can use the microphone (see §5).
Chrome is made by Google, and Google’s own privacy policy applies to Chrome.

3.What Saypen does not do

No network requests of any kind: no servers, no uploads, no cloud processing.
No recording, saving or sending of audio.
No history, no transcripts and no log of what you dictate.
No analytics, no ads and no tracking.
No accounts and no sign-in.
No remote code: everything Saypen runs is inside the extension package.
No reading of other page content, cookies, form data or browsing history.
No sale, transfer or sharing of your data with anyone, and no human access to it.
No use of your data for advertising, creditworthiness, lending or any other purpose unrelated to voice typing.

4.Storage and retention

What	Where	How long
Microphone audio	Nowhere. It goes to Chrome’s on-device recognizer and is never stored by Saypen.	Not kept
The few characters before the cursor	Nowhere. They are read when a phrase is typed.	Not kept
The current voice-typing session	`chrome.storage.session`: memory only, never written to disk	Until voice typing stops. Also cleared when the browser starts.
Text that could not be typed	Saypen’s bar on that page	Until you copy or close it
Settings	`chrome.storage.local`, on this computer only	Until you change them, or until you remove Saypen from Chrome
Speech model for a language	Chrome’s own storage	Managed by Chrome, not by Saypen

Text you dictate into a website becomes part of that page, as if you had typed it yourself. From then on, that website handles it under its own terms and privacy policy.

5.Microphone

Chrome asks for microphone permission once, on Saypen’s own setup page, not on every website. When Chrome asks, choose “Allow while visiting the site”.
The microphone is open only while voice typing is on, or while the “Try it here” box on the setup page is listening. Saypen’s toolbar icon shows a red “ON” badge while voice typing is on, and your computer may also show that the microphone is in use.
Voice typing stops when you press Esc, click Stop in Saypen’s bar, press the shortcut or click the toolbar button again, or say “stop listening”. It also stops on its own after a silence (1 minute by default; you can choose 30 seconds, 2 minutes, 5 minutes or never), when you switch tabs, when you switch to another window or app (after about 1.5 seconds), when the page navigates, and when the tab closes.
Saypen is not a meeting transcriber. It does not transcribe tab audio, and it only listens to your microphone while you have voice typing on.
To withdraw the permission, open chrome://settings/content/microphone. Saypen is listed there by the address of its setup page, which starts with chrome-extension://. Change it to Block, or remove it. Voice typing then cannot start until you allow the microphone again on the setup page.

6.Access to web pages

Saypen has no access to any website in advance: it requests no host permissions.
It gets temporary access to the current tab only when you press its shortcut (Alt+Shift+D by default) or click its toolbar button. This is Chrome’s activeTab permission. The access ends when you leave the page.
In that tab, Saypen runs only its own bundled script, which finds the focused text field, types the recognized text, carries out the voice commands, shows the listening bar and listens for Esc.
Saypen never types into password fields.
Chrome doesn’t let extensions run on its own pages (such as chrome:// pages), on the Chrome Web Store or in its PDF viewer, so Saypen cannot type there.
Incognito: like any extension, Saypen runs in incognito windows only if you allow it on its details page in Chrome. The same rules apply there.

7.Permissions and why Saypen needs them

Permission	What it is used for
`activeTab`	Temporary access to the current tab, only after you press the shortcut or click the toolbar button. Saypen can’t see or touch any site until then, and the access ends when you leave the page.
`scripting`	Runs Saypen’s own bundled script in that tab’s frames: it finds the focused text field, types the recognized text, handles the voice commands (line breaks, deleting the last phrase), shows the listening bar and listens for Esc.
`offscreen`	Chrome’s background script for extensions can’t use the microphone or the speech recognizer. A hidden page of Saypen’s own holds the microphone and Chrome’s on-device recognizer, only while voice typing is on, and is closed when voice typing stops.
`storage`	Saves your settings on this computer (`chrome.storage.local`) and keeps the state of the running voice-typing session in memory (`chrome.storage.session`), as described in §1.

These four permissions show no warning when you install Saypen. The microphone is not requested at install: Chrome asks for it the first time, on Saypen’s setup page.

8.Security

Everything Saypen keeps stays in your Chrome profile on your computer, protected by your operating-system account like the rest of your browser data. Before you dictate something private, check that the cursor is in the field you mean. Don’t dictate passwords: the words you speak are shown in Saypen’s bar, which is why Saypen never types into password fields.

9.Children

Saypen is a general-purpose tool. It isn’t directed at children, and it doesn’t collect personal information from anyone.

10.Chrome Web Store User Data Policy and Limited Use

The use of information received from Google APIs will adhere to the Chrome Web Store User Data Policy, including the Limited Use requirements. Saypen uses the data it handles only to provide its single purpose described above, keeps it on your computer, and never sells it, transfers it or uses it for any other purpose.

11.Changes to this policy

If Saypen’s data practices change, this page will be updated, with a new effective date, before the new version of the extension is published, and the changes will be summarized at the top of this page. If a change matters for your privacy, Saypen will also tell you about it inside the extension before it takes effect.

12.Contact

Questions or requests about this policy:

endritn32@gmail.com (Iliria Studio)

The English version of this policy is the authoritative one.