Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/pdf-math-translate-guide
Translate scientific PDFs with preserved math formatting via PDFMathTranslate
Passed 22 of 22 test cases with the skill, and 14 of 22 without it.
and 7 more in the full report
8 fixed14 passed either way0 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
I prefer using a graphical browser interface to upload academic papers rather than command line flags. What single flag passed to pdf2zh launches the web-based Gradio interface?
| Checked | Without skill | With skill |
|---|---|---|
| The command uses pdf2zh -g | Missed | Passed |
The instructions your model reads when the skill loads.
PDFMathTranslate is an open-source tool designed specifically for translating scientific and technical PDF documents while preserving mathematical formulas, tables, figures, and the overall layout structure. Traditional PDF translators often mangle equations and destroy formatting, making translated papers difficult to read. PDFMathTranslate solves this problem by intelligently detecting and preserving mathematical content during the translation process.
The tool leverages large language models for high-quality translation while maintaining the integrity of LaTeX-rendered equations, chemical formulas, and complex table structures commonly found in academic publications. It supports translation between dozens of language pairs, making it invaluable for researchers who need to read papers published in languages outside their expertise.
PDFMathTranslate has gained significant traction in the academic community with over 32,000 GitHub stars, reflecting the widespread need for reliable scientific document translation that respects the specialized formatting requirements of research papers.
Install PDFMathTranslate using pip in a Python environment (Python 3.8 or higher required):
pip install pdf2zh
For GPU-accelerated processing, install with CUDA support:
pip install pdf2zh[cuda]
Configure your translation backend by setting the appropriate environment variable. The tool supports multiple LLM providers:
# Using OpenAI-compatible APIs
export OPENAI_API_KEY=$OPENAI_API_KEY
export OPENAI_BASE_URL=$OPENAI_BASE_URL
# Using Google Translate (free, no key required)
pdf2zh input.pdf -s google
# Using DeepL
export DEEPL_AUTH_KEY=$DEEPL_AUTH_KEY
pdf2zh input.pdf -s deepl
You can also launch the web-based GUI for interactive use:
test cases passed, vs 14 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.