Skip to content

RIS, EndNote, and PubMed NBIB to BibTeX converter

Transform Zotero, Mendeley, EndNote, and PubMed exports into clean, deduplicated BibTeX keys with strict LaTeX escaping and round-trip export.

Your files are open only in this tab. Nothing is uploaded.

RIS, EndNote, PubMed NBIB, or BibTeX input

Normalizes citation keys, escapes LaTeX symbols, deduplicates DOIs, and exports .bib, .ris, .enw, CSL-JSON, or Excel.

Drop a .ris, .enw, .nbib, .bib, or .json citation file

Manuscript references are processed offline on your device.

How to use RIS & EndNote to BibTeX

  1. Load or paste your reference file. Drop a .ris, .enw, .nbib, .medline, .bib, or .json export, paste raw citation records into Or paste citation records, or click Load sample RIS to inspect four built-in entries.
  2. Select citation key formatting and cleanup rules. Choose authorYear, author_year_word, or authorYearTitle under Citation key format and toggle Escape LaTeX characters (&, %, _) or Deduplicate by DOI / title.
  3. Click Convert citations and filter records. Review the detected input format, duplicate removal count, and search inside Normalized records by cite key, author name, paper title, or DOI string.
  4. Copy or export the cleaned bibliography. Click Copy BibTeX for Overleaf or choose Download .bib, Download .ris, Download .enw, Download CSL-JSON, or Download Excel.

Your exam questions, student work and manuscripts stay in this browser tab on your device. Nothing is uploaded or logged.

Parsing RIS, EndNote .enw, and PubMed NBIB files locally

Reference conversion runs inside edu.worker.ts without transmitting unpublished literature reviews, systematic-review screenings, or grant bibliographies over external networks. Click Load sample RIS to test sample-systematic-review.ris, or drop .ris, .enw, .nbib, .medline, .bib, and .json files directly into the workspace.

Auto-detecting tagged records and continuation lines

Before inspecting tags, parseMixedCitations strips any leading 3 bytes of a UTF-8 byte order mark (0xEF 0xBB 0xBF) commonly added by Windows exports. It checks file extensions alongside structural markers such as TY - for RIS, %0 for EndNote, @article{ for BibTeX, or PMID - for PubMed exports.

MEDLINE .nbib records follow the NIH NCBI PubMed Help manual structure. The parser unfolds wrapped titles indented with 2 or more spaces, prefers full FAU author names over AU initials, extracts 4-digit years from DP, turns single hyphens in PG page spans into -- ranges, and parses [doi] values from LID or AID lines.

Keeping unpublished manuscript references inside browser memory

Academic screening lists frequently reveal active research hypotheses and pre-publication peer review selections before journal release. Because every parsing and normalization pass runs inside client-side browser tabs with 0 bytes sent to remote endpoints, confidential literature exports remain on your workstation.

You can paste multi-record text straight into Or paste citation records or upload full reference library downloads without opening desktop citation managers. When you need formatted reference lists from cleaned BibTeX later, format bibliography entries into APA, MLA, or IEEE styles inside the same offline workspace.

How does LaTeX character escaping prevent BibTeX compile errors?

Unescaped ampersands (&), percent signs (%), $, #, and underscores (_) inside journal names or titles trigger immediate TeX math-mode, comment, or alignment errors during pdflatex compilation. Enabling Escape LaTeX characters (&, %, _) converts reserved punctuation and accented letters into safe macro equivalents before writing .bib fields.

Escaping reserved symbols and European diacritics safely

Following the CTAN BibTeX bibliography package documentation, escapeLatexField converts &, %, $, #, and _ into backslash escapes while translating en-dashes to -- and em-dashes to ---. European diacritics including ü, ä, ö, é, è, á, ó, ñ, ç, and ß convert cleanly into braced TeX accent sequences.

Existing backslash sequences such as \& or \' pass through untouched so previously escaped BibTeX files never get double-escaped. You can also catch missing citations before journal submission once your manuscript text and cleaned .bib file match.

Generating collision-free author and year citation keys

Under Citation key format, choose authorYear (turing1936), author_year_word (turing_1936_computable), or authorYearTitle (turing1936Computable) citation key patterns. Key builders normalize Unicode accents to ASCII, strip non-alphanumeric punctuation, and ignore 18 common English stop words (a, an, the, for, with, and others) when picking the title token.

If multiple publications share the same author and year, collision counters append lowercase letters a through z across the first 26 duplicates and numeric suffixes thereafter. Deterministic order ensures repeated conversions yield identical reference keys when collaborating across co-author drafts.

Deduplicating DOIs and exporting across bibliography formats

Literature searches across PubMed, Scopus, Web of Science, and Zotero routinely return overlapping citations with mismatched URL prefixes. Checking Deduplicate by DOI / title strips doi.org/ and doi: resolver prefixes into lowercase strings matching the DOI Foundation identifier specification, collapsing matching records automatically.

Export button File extension Primary target apps Page & dash normalization Key fields included
Download .bib .bib LaTeX, Overleaf, JabRef -- range delimiters Escaped authors, title, journal, DOI
Download .ris .ris Zotero, Mendeley, Covidence Split SP / EP tags TY, ID, AU, TI, JO, PY, DO
Download .enw .enw Clarivate EndNote desktop Single - hyphen ranges %0, %A, %T, %J, %D, %R, %F
Download CSL-JSON .json Pandoc, Quarto, Hugo Preserved page string id, type, author[], issued, DOI
Download Excel .xlsx Screening sheets, PRISMA logs Plain text column cell 10-column References worksheet

Records lacking a DOI still deduplicate whenever normalized alphanumeric titles (6 characters or longer) and publication years match identically. Medical or graduate students turning systematic-review flashcards into study material can also open an offline Anki flashcard deck browser without external accounts.

Tip: Load Load sample RIS (4 records including a duplicated Vaswani 2017 DOI and Clara Müller) to verify deduplication (1 duplicate removed) and LaTeX diacritic escaping before processing a multi-hundred-record PubMed .nbib batch.

Filtering normalized reference tables and round-trip audits

After clicking Convert citations, the Normalized records grid lists each computed Cite key, lead author family name, publication year, and full title. Typing into Filter key, author, title, or DOI... filters matches live across every normalized field before copying or exporting.

Checking missing years and author arrays before compilation

Behind the scenes, inspectRisBibtex checks normalized entries and caps diagnostic warnings at 12 entries whenever publication years or author arrays are blank. Reviewing those gaps in Normalized records catches incomplete webpage exports or n.d. placeholders before running bibtex or exporting to Citation Style Language CSL specification CSL-JSON pipelines.

Clicking Copy BibTeX writes the 14-row readonly preview directly to your system clipboard and toggles the confirmation badge for 1800 ms before resetting. Keeping both instant clipboard snippets and downloadable RFC 4180 compatible spreadsheets makes multi-author review cycles straightforward.

Pairing systematic literature reviews with private archives

Clinical investigators managing systematic reviews alongside anonymized case series frequently keep both bibliography exports and DICOM imaging files strictly local. When preparing radiology teaching files or imaging appendices for the same paper, inspect and scrub study tags in our DICOM Viewer & Anonymizer inside browser memory without external uploads.

Questions

Which citation formats does the converter auto-detect?

The parser detects PubMed NBIB exports by PMID and TI lines, RIS files by TY - tags, EndNote archives by %0 markers, BibTeX entries by @type headers, and CSL-JSON arrays or objects. It automatically strips leading 3-byte UTF-8 byte order marks before reading the first line.

Why does LaTeX fail on unescaped ampersands and underscores in BibTeX?

TeX treats unescaped ampersands as table column delimiters, percent signs as end-of-line comments, hashes as macro parameter tokens, and underscores as math-mode subscript operators. Escaping reserved punctuation into backslash sequences inside title and journal fields prevents pdflatex compilation halts.

How does DOI and title deduplication merge overlapping exports?

When Deduplicate by DOI / title is enabled, the worker normalizes DOIs by stripping doi.org resolver prefixes and lowercase doi: tags before comparing records. Citations lacking a DOI still collapse cleanly whenever normalized alphanumeric titles of at least 6 characters match the same publication year.

How are duplicate citation keys like smith2024 resolved automatically?

Whenever two normalized entries produce an identical base key, a collision tracker appends lowercase letters a, b, c, and beyond sequentially across the entire batch. After exhausting 26 letters, numeric suffixes keep every citation key strictly unique inside Overleaf and local BibTeX builds.

Does Escaping LaTeX characters double-escape existing backslash commands?

No. The character scanner inspects each backslash pair inside author, title, and journal fields and copies existing sequences such as \& or \' verbatim before checking plain metacharacters and European accented letters including ü, ö, ä, é, ñ, and ß.

Why does PubMed NBIB import prefer FAU tags over AU tags?

MEDLINE NBIB records store complete author given names and surnames in FAU lines while AU lines hold abbreviated surname initials. Preferring FAU whenever present preserves complete author names when building BibTeX author fields, RIS AU lines, and CSL-JSON author arrays.

Updated