Textuary — Article Reader & Text Sanctuary

Textuary transforms the article in your current tab into a calm, distraction-free text sanctuary. Features: - clean article extraction without publisher-specific rules - estimated reading time, live progress and time remaining - four typefaces plus adjustable text size, spacing and column width - paper, evening and automatic ambient themes - locally remembered reading preferences - choose system voices or optional natural Kokoro voices generated locally with WebGPU - selectable voice and playback speed - pause, resume and stop controls - automatic highlighting of the passage being read aloud - an account-free local Library with offline clean-text snapshots - Library search, read/unread controls and restored reading and read-aloud positions - same-tab return from the Library without duplicate article tabs - clean printing and PDF output using the selected reader typeface - a keyboard shortcut for quick access Textuary runs only when you activate it. Article extraction and sanitisation happen locally in the browser. Reading and speech preferences are saved in local browser extension storage. An article snapshot, its original URL and reading state are stored only when the user explicitly chooses Save article; Textuary does not upload or synchronise the Library. If the user enables Natural voices, the browser downloads and caches approximately 330 MB of model data from Hugging Face; the packaged runtime still generates speech locally and never transmits article text.Textuary contains no advertising or analytics and does not transmit article text or browsing activity to the developer. Some pages do not deliver their complete article to the browser. Textuary can reorganise content that is available in the page, but it cannot retrieve text that was never delivered.