Xpeaker — Privacy Policy

Last updated: 2026-07-13

Xpeaker does not collect, store, transmit, or sell any personal information. It runs entirely on your device.

• No analytics, tracking, advertising, or account — and none of your data leaves your device. The only network request Xpeaker makes is optional: if you enable the Mood ring (Settings), it downloads a small emotion-classification model once from HuggingFace (huggingface.co) and caches it. Classification then runs entirely on your device — the text of your posts is never sent to HuggingFace or anywhere else. If you don't enable the mood ring, Xpeaker makes no network requests of its own.
• Settings stay local. Your preferences (voice, speed, per-author voices, keymap, etc.) are saved via chrome.storage.local and never leave your device.
• Post text is used only on your device. The text of a post you choose to read is taken from the page and handed to your browser's text-to-speech engine (chrome.tts) — and, if installed, the separate Supertonic Text-to-Speech Voices extension — solely to generate audio on your device; if you enable the mood ring, it is also classified for emotion on your device. It is not logged, stored, or transmitted by Xpeaker.
• Background scenes read only what's already on the page. For a ticker shader mode, Xpeaker hovers a post's $cashtag to make X show its own price-card popover, then reads the ticker, % change, and logo from it and renders them in the shader — all in your browser; Xpeaker transmits none of it. (That hover may prompt X — not Xpeaker — to fetch the card from X's own servers, exactly as it would if you moused over the cashtag yourself.)

Permissions

• tts — speak the text aloud.
• storage — save your settings locally.
• contextMenus — add Xpeaker entries to the right-click menu on X/Twitter.
• offscreen — run the on-device emotion classifier (mood ring) in a hidden document.
• Access to huggingface.co — download the mood-ring model once, only if you enable it.
• Content script on x.com / twitter.com — show the read-aloud buttons, player bar, and Focus overlay.

If you install the optional Supertonic Text-to-Speech Voices companion extension, its own privacy policy governs how it synthesizes audio; Xpeaker only requests speech from it locally via the browser's TTS API.

Not affiliated with X Corp or Twitter, Inc. "X" and "Twitter" are referenced only to describe the website Xpeaker works on.

Questions: github.com/dgnsrekt/xpeaker/issues