sumatrapdf

Application

sumatrapdfreader/sumatrapdf

A multi-format document reader for Windows supporting PDF, EPUB, MOBI, and more.

Overview

SumatraPDF is a fast, open-source document reader for Windows. It supports PDF, EPUB, MOBI, CBZ, CBR, FB2, CHM, XPS, and DjVu formats. The project is licensed under GPLv3 and has a lightweight, portable design.

README Preview

[](https://github.com/sumatrapdfreader/sumatrapdf/actions/workflows/build.yml)
## SumatraPDF Reader

SumatraPDF is a multi-format (PDF, EPUB, MOBI, CBZ, CBR, FB2, CHM, XPS, DjVu) reader
for Windows under (A)GPLv3 license, with some code under BSD license (see
AUTHORS).

More Information:
* [Website](https://www.sumatrapdfreader.org/free-pdf-reader)
* [Manual](https://www.sumatrapdfreader.org/manual)
* [Developer Information](https://www.sumatrapdfreader.org/docs/Contribute-to-SumatraPDF)

FAQ (3)

Troubleshooting
How to fix SumatraPDF printing the wrong page when entering a custom page range (e.g. typing 2 prints page 3)?

This incorrect page printing bug affects older SumatraPDF versions such as 3.5.2: inputting values like 2, 2-2, or 1 as the custom page range always prints page 3. The issue was addressed in PR #5108, is confirmed fixed in SumatraPDF 3.6, and is also resolved in the 3.7 pre-release. Upgrade your SumatraPDF installation to version 3.6 or later to fix the problem.

GitHub Issue #5623
Troubleshooting
Why does SumatraPDF -search or -dde "[Search]" silently ignore non-ASCII (e.g. Chinese) search terms when an instance is already running?

This is a confirmed bug in SumatraPDF (observed in pre-release 3.7.0): when the search command is forwarded to an already-running instance via -search or -dde "[Search(...)]", the non-ASCII search term gets corrupted internally (likely converted through a narrow/ANSI string in the Search handler), so no find is executed and no highlight or jump occurs — even though the DDE transport itself delivers the Unicode text intact.

Workaround: make sure no SumatraPDF instance is running before issuing the command, so the new process executes -search in-process:

taskkill /IM SumatraPDF.exe /F
SumatraPDF.exe "D:\file.pdf" -search "火"

When launched this way, the term is correctly found and highlighted (debug log shows TaskFindCount and ShowMatchCount). The downside is losing the current window session, which can be partially recovered via session restore (Settings > Advanced Options > RestoreSession = true). ASCII-only search terms and other DDE commands like [GotoPage] are unaffected.

GitHub Issue #6412
Troubleshooting
Why does SumatraPDF 3.6 freeze or block page navigation while a page is rendering?

This is a regression introduced in SumatraPDF 3.6: slow-rendering pages block the UI, so pressing the right arrow key or dragging the scroll bar has no effect until rendering finishes, and the displayed page lags behind the scroll position. Workaround: download the latest pre-release build (3.7+) from https://www.sumatrapdfreader.org/prerelease — it changes rendering so navigation is no longer blocked. Note: with very slow PDFs the main view may still lag slightly behind the scrollbar; this residual lag is a known issue. Alternatively, downgrade to version 3.5.x, where scrolling and page display stayed in sync.

GitHub Issue #6362