PDF & images

Reference material lives next to your notes. Link a PDF with a wiki-link or drop one onto a note and it opens in a dedicated viewer tab; paste or drag an image into a note and it renders inline.

PDFs

Link a PDF with [[file.pdf]] or ![](file.pdf), or drop one onto a note, to open it in a dedicated viewer tab.

A viewer that scales

The viewer is page-virtualized: only pages near the viewport are rasterized, and scrolled-away pages free both their memory and their GPU texture. An 800-page document stays light. Rendering is DPI-aware, so pages stay crisp on high-resolution displays.

Zoom, page navigation, and a table of contents for jumping around a long document. Two fit modes live next to the zoom controls: ↔ fits the page width to the viewer, ⤢ fits the whole page — both stay in effect as you resize the window, until you zoom manually.

Highlight markup

Drag to highlight a passage in a PDF. Each highlight becomes a markup with note↔PDF jump links: follow it from your note to land on the right page, or from the PDF back to where you wrote about it.

Next to the highlight pen sits the area tool (⬚): drag a box over any region — a figure, a chart, a scanned page with no text layer at all — and it becomes a markup exactly like a text highlight, same colors and jump links.

Password-protected PDFs

Encrypted PDFs are supported — both RC4 and AES. Zorite prompts for the password and renders the document once it’s unlocked. On the rare file Zorite can’t parse at all (an exotic encryption scheme, unusual features), the viewer offers Open in system viewer to hand it to your OS default app instead.

Fill in forms

A PDF with form fields (AcroForm) is fillable right in the viewer. Click a checkbox or radio button to toggle it. Click a text field and a small input appears seated under it — type, then Enter (or click away) saves, Esc cancels, and Tab / Shift-Tab hop field to field, scrolling each into view. Edits are written back into the stored PDF with proper appearance streams, so the filled file renders correctly in every other PDF viewer too. Signature and read-only fields stay untouched. Choice fields (dropdowns and lists) offer their options as a click-to-pick list under the input — typing still works for combos that allow free entry.

Under the hood

PDF rendering is pure-Rust via hayro, so there are no native dependencies. The viewer is its own reusable gpui-pdf crate.

Inline images

![](path-or-url) images render for real in your notes. To add one, paste or drag-and-drop a file — it’s copied into the data directory’s images/ folder, so the note doesn’t depend on where the original lived. Importing the same image twice (paste or drag, any page) reuses the existing copy instead of duplicating it, and Settings → General → Unused images sweeps files nothing references anymore into the system trash (with a confirmation listing them first).

Drag the corner handle to resize an image; the new size is saved back into the Markdown as {width=N}, so it stays the size you set the next time the note renders.

Export a note to PDF

Right-click a tab or a sidebar page and pick Export as PDF… (or press ⌘P / Ctrl+P on the active note). The export renders like the reading view — wrapped styled text, tables, inline images, alerts, typeset math, and mermaid diagrams — written by a pure-Rust PDF writer, no browser or print dialog involved.

One known limitation: the writer places glyphs directly without a text-shaping pass, so font ligatures (the ->/=> arrows of a font like JetBrains Mono, or a serif’s fi) render in the app but come out as their individual characters in the exported PDF. The text is unaffected — only the glyph substitution is skipped.