ArcaRead — Privacy Policy

Last updated: April 2026
Short version: ArcaRead does not transmit any data off your device. There is no server operated by the author. No analytics, no crash reporter, no telemetry. The full source code is public.

What data ArcaRead accesses

To apply the bionic-reading effect, the extension runs a content script on every page you visit. That script reads text nodes in the DOM so it can bold the first half of each word. It does not transmit any of that text, to the author or to anyone else.

What data ArcaRead stores, and where

AI summaries

The Summarize button in Reader uses Chrome's built-in on-device Gemini Nano model. Your article text is passed to the model locally; nothing leaves your computer. If on-device AI is not available on your browser or machine, the button shows a clear "not available" message — there is no network fallback and no data is sent to any remote AI service.

PDF handling

When you opt in to Open PDFs in reader, the extension fetches the PDF bytes from the URL you navigated to and renders them locally via PDF.js. The PDF fetch comes from your browser, to the PDF's original host. No third party sees the fetch.

Permissions, and why

Third-party services

None. ArcaRead does not integrate any third-party analytics, telemetry, advertising, or crash-reporting service.

Children's privacy

ArcaRead is a reading-aid tool with no account, no login, and no data collection. It is suitable for use by children subject to their parent's or guardian's judgement about web use in general.

Changes to this policy

If this policy changes, the updated version will be committed to the source repository and the "Last updated" date at the top of this page will change. No mechanism exists for us to notify you individually because we do not have any way to reach you.

Contact

Questions, concerns, or bug reports: open an issue on the GitHub repository.