google-research/arxiv-latex-cleaner — explained in plain English
Analysis updated 2026-06-24
Clean a LaTeX paper folder before uploading to arXiv to automatically remove comments and unused files
Strip all todo and author shorthand commands from a LaTeX paper before making the source public
Resize and convert images in a LaTeX project to fit within arXiv's 50 MB size limit
Replace TikZ diagram source code with compiled image files to keep your diagram code private on arXiv
| google-research/arxiv-latex-cleaner | xfangfang/macast | yihong0618/xiaogpt | |
|---|---|---|---|
| Stars | 6,838 | 6,837 | 6,837 |
| Language | Python | Python | Python |
| Setup difficulty | easy | easy | moderate |
| Complexity | 2/5 | 2/5 | 3/5 |
| Audience | researcher | general | general |
Figures from each repo's GitHub metadata at analysis time.
This is a command-line tool for researchers who write academic papers in LaTeX and need to submit them to arXiv, the popular preprint server. The tool takes your paper's source folder and produces a cleaned-up copy that is ready to zip and upload. The cleaning process handles several practical concerns. On the privacy side, it strips all comments from your LaTeX code, which would otherwise be visible to anyone who downloads the source from arXiv. It can also remove helper commands like todo notes that you defined during writing but do not want in the final submission. On the file size side, arXiv has a 50MB limit, so the tool removes unused image files and unused tex files, and can optionally resize images, compress PDFs, or convert PNG files to JPG to shrink the package. For diagrams drawn with TikZ (a common way to create figures directly in LaTeX), the tool can replace the raw source code with pre-compiled image files, which hides your diagram code from public view. There is also support for custom find-and-replace rules written in a config file, which lets you swap out your own shorthand commands for standard LaTeX before submitting. Installation is straightforward via pip or Homebrew on Mac. You point it at your input folder, optionally pass flags or a config file, and it produces a new cleaned folder alongside the original. The project is from Google Research and is written in Python, requiring version 3.9 or higher.
A command-line tool from Google Research that prepares a LaTeX paper for arXiv by stripping comments, removing todo notes, cutting unused files, and shrinking images to fit under the 50 MB upload limit.
Mainly Python. The stack also includes Python, LaTeX.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly researcher.
This repo across BitVibe Labs
Verify against the repo before relying on details.