Read Zotero PDFsas source-linked Markdown

Reflow columns, formulas, tables, and figures — then jump back to the original PDF.

Built for people who still read papers in Zotero

  • Zotero 7–10
  • MIT licensed
  • Mktero account
  • Local cache

Every excerpt has a way back

Text, formulas, tables, and figures stay mapped to PDF pages and regions. If the match is uncertain, Mktero does not guess.

Mktero reading view with continuous text, outline, and a table

One loop for papers that are hard to read

Convert once

The default Mktero account reflows two-column layouts into continuous Markdown. No separate OCR key is required. The same PDF is cached locally.

Annotate in place

Highlights, underlines, and comments stay in the reading tab and sync to Zotero PDF annotations.

Verify the source

Jump from a sentence to a reliable page and region, or copy it with paper, page, and Zotero link.

Keep it local

Credentials, Markdown, figures, and maps live in your Zotero profile. Mktero does not sync them.

Built for papers that fight the reader

Open from Zotero, read as continuous text, and keep structure, figures, and notes in one tab.

Mktero converting a Zotero PDF into a Markdown reading view

Start from the PDF

A local file becomes a reading-first Markdown tab. A cache hit skips the remote conversion.

Creating a Zotero-synced annotation in the Mktero reading tab

Notes stay beside the sentence

Create, recolor, comment, or delete annotations without leaving the line you are on.

Correcting recognized text in the continuous reading view

Fix what OCR missed

Edit a paragraph or table cell without rewriting the immutable conversion. Restore anytime.

It lives in the tools you already use

Mktero is a Zotero extension. The default conversion uses your Mktero account. Switch to your own OCR if you do not want the PDF sent to Mktero.

  1. Zotero 7–10

    A restartless reading tab: outline, citations, annotations, and the library you already keep.

  2. MinerU

    Switch to your own MinerU cloud or a local MinerU service if you do not want the PDF sent to Mktero. This is not the default path.

  3. Mistral OCR

    Another alternative. With your own Mistral key, the PDF is not sent to Mktero.

From download to first page

About five minutes, and the default path needs no separate OCR key.

  1. Download the XPI

    Take the latest release file. Zotero then keeps formal releases up to date on its own.

  2. Add it in Zotero

    Choose Tools → Plugins, open the gear menu, and pick Install Add-on From File.... No restart is needed.

  3. Sign in and read

    Sign in under Settings → Mktero, then right-click a PDF and choose Read as Markdown with Mktero.

Translation, and references you already have

Optional AI translation never rewrites the source. The citation graph only links items already in the current library.

Selecting text in Mktero for translation or citation actions

Original, translation, or bilingual

Translation stays opt-in. Switch views without losing your place. The source Markdown does not change.

See papers already in your library

Match direct references by DOI or arXiv, refreshing from Semantic Scholar, OpenCitations, and OpenAlex when needed.

  • Export corrected Markdown and figures
  • Save a portable Zotero snapshot
  • Import references from the citation popup

A reading-first Zotero tab

Outline, continuous text, figures, and notes in one place. The PDF stays the source of truth.

Mktero reading tab with outline and body on one screen

Know where the PDF goes

With the default converter, a cache miss sends the complete PDF to the Mktero service you signed in to. If you switch to your own MinerU or Mistral, the PDF goes there instead and is not sent to Mktero. Everything else stays in the local profile.

Read the privacy notes

Keep the paper in Zotero. Keep reading.

Install the XPI, sign in to Mktero, and open the next PDF as Markdown.

Mktero account for the default path · Zotero 7–10 · MIT License