SAPIENTIA Research

SAPIENTIA is a tool to search, read, and listen to your PDFs - from papers and
dictionaries to everyday reading. Find with cross-PDF search, read in the reader, and
listen with on-device text-to-speech. The AI exploration that breaks research silos is here too.

[Instant search across thousands of pages]

Import a PDF and full-text fuzzy search becomes available.
Hiragana, katakana, full and half width are unified; Levenshtein distance allows for typos.
Even a dictionary of thousands of pages returns results almost instantly.

Two-phase (exact then fuzzy), Japanese normalization, context snippets, seven highlight styles.

[Peripheral Exploration]

AI extracts key phrases and generates cards in two ranges: Mid-range (adjacent fields)
and Far-range (distant domains, including folk knowledge and crafts across cultures).
Paper lookup runs in Japanese, English, and Chinese in parallel against Semantic Scholar,
so even a Japanese keyword pulls in the right papers; otherwise it falls back to Google
Scholar or Web search.

[Multilingual Paper Search]

A keyword in any language is auto-translated and searched across CiNii, J-STAGE,
Crossref, arXiv, and Semantic Scholar at once. Result titles are shown in your UI language.

[AI Summary - on-device when possible]

Summarize the context around a hit, or all hits at once. On Apple Intelligence devices
(iPhone 15 Pro and later) it runs fully on-device - PDF content never leaves your device.
Otherwise it falls back to the Gemini API.

[Write memos and return to the source]

Leave a memo on any page. v4.0 introduces a dedicated Memos tab
where every memo can be browsed and searched together.

One tap returns you to the source PDF and page.

[Read - resume, bookmarks, cross-PDF hits (v4.1+)]

The app remembers the page (and read-aloud position) you closed on, so you can resume.
Bookmark and jump from the bottom bar, and move through pages with the slider. After
searching, step through exact-match hits one by one - across other PDFs too - each
auto-highlighted.

[Listen - on-device text-to-speech (v4.2)]

Have your pages read aloud, continuing automatically to the next page. Adjustable speed,
sleep timer (up to 2 hours), pages-to-read limit (up to 100), resume mid-sentence, an
"extend" prompt when time/pages run out (tap or shake to continue), and a remaining
countdown. Plays in the background and on the lock screen, with controls and extend.
Runs on-device; text is never sent externally. Tip: add a high-quality voice
in iOS Settings for a more natural sound.

[Capture pages directly]

No PDF? Photograph book pages and we OCR them via Apple Live Text,
including vertical Japanese. Capture an entire dictionary or specialized book.

[All major cloud providers]

Import PDFs from iCloud Drive, Google Drive, OneDrive, and Dropbox.

[Multilingual UI and translation]

UI in Japanese, English, and Chinese. Tap to translate selected text; on Apple
Intelligence devices, translation runs on-device too.

[Pro features (14-day free trial)]

The first 14 days unlock every feature for you to try.
After that, Pro features can be permanently unlocked with a one-time purchase.

Always free: unlimited PDF import & search, basic highlighting, memos, up to 3 folders, history.

Pro features (one-time unlock): Peripheral Exploration, unlimited AI Summary (on-device when
supported), multilingual paper search across 5 databases, translation, continuous-capture OCR,
unlimited folders, no daily caps. v3.0 buyers keep Pro automatically.

[Who is this for?]

Researchers and students; faculty and postdocs; independent researchers; translators and
editors who consult dictionaries; anyone managing many PDFs.

[Notes on use]

AI summaries are generated by Google Gemini 2.5 Flash Lite or Apple Foundation Models;
always check the original before citing. Academic results come via each service's public API.
PDF data is processed on-device by default (no external transmission on Apple Intelligence devices).