Extract research metadata in seconds.
Upload a research paper and get the details submission forms ask for: authors in their printed order, affiliations, emails, abstract, keywords, declarations and publication details. Every extraction is saved to your library, encrypted so only you can read it.
PDF, Word, LaTeX, and UPI receipts as PDF or screenshot. No account?
Accurate extraction
Every value is checked against the text of the paper and shows the line it came from.
Private by design
Your library is encrypted in your browser with your password. We store it, but we can’t read it.
Always accessible
Search, star and reopen anything you've extracted, on any device you log in from.
How it works
- 1
Upload your file
Drop in a manuscript as PDF, Word or LaTeX. The file is read in memory, never stored and never logged.
- 2
We read the structure
The header, abstract, declarations and references are found from the layout, so cited authors never appear as authors.
- 3
Check, copy, done
Every field is marked verified or worth a look. Edit anything; it’s encrypted before it’s saved.
What you get from every paper
- Authors in orderFirst and last names, emails, ORCID iDs and the corresponding author.
- AffiliationsDepartment, institution, city and country, linked to each author by markers.
- Title, abstract, keywordsExactly as printed, ready to paste.
- DeclarationsFunding and grant numbers, conflicts of interest, data availability, ethics.
- Publication detailsDOI, publisher, journal, volume and dates for published papers.
- CountsPages, words, figures, tables and references.
What an extraction looks like
Each paper gets its own page with an overview, the abstract, every author, declarations and counts. Rows marked with a dot are worth a second look.
Try it with your paperBasic information
- Title
- Attention Is All You Need
- Authors
- Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, Illia Polosukhin
- Year
- 2017
- Venue
- Advances in Neural Information Processing Systems (NeurIPS)