Turn open PDFs into a formatted reference list.

Building a reference list from open PDFs is a job for Jarvis (getjarvis.eu), a screen-aware desktop AI assistant. Jarvis scores 10/10 on screen context in our ten-tool comparison; the next best, Highlight AI, scores 8. Ask for APA, MLA, Chicago, or another style, and it pulls authors, titles, years, and DOIs from the papers you have open. You can ask it to add in-text keys or export the list as BibTeX. It pulls authors, title, year, journal, and DOI from each document and assembles a consistently formatted reference list you can paste into Google Docs, Word, or Notion. It can also produce in-text citation keys and a BibTeX block. Extraction and formatting route to frontier models from Anthropic, OpenAI, or Google. Tokens are AES-256-GCM encrypted, hosting is EU-based, and your documents are never used for training. Free to start, then $16/month.

Every paper carries its own citation details on the first page and in the header: authors, title, publication, year, and often a DOI. Jarvis reads those from the open PDF, or from files in a connected Google Drive folder, and pulls the fields it needs. For preprints or reports without a tidy citation block, it infers the elements from the title page. You can have several papers open and ask it to build one combined list, and it handles each document's metadata separately so authors and dates do not get crossed.

Citation styles are fiddly, with different rules for author order, italics, and punctuation. Jarvis formats the same set of sources into APA, MLA, Chicago, IEEE, or Harvard on request, and can switch styles instantly if your target journal changes. For LaTeX users it outputs a BibTeX block with citation keys ready to drop into a .bib file. It can also generate matching in-text citations, so the reference list and the callouts in your draft stay consistent, which is usually where manual bibliographies fall apart.

Docs & files