Citation Methodology

What NeatCite does

NeatCite turns structured source metadata into a reference-list entry and an in-text citation. You can paste a DOI, PubMed ID, ISBN, or URL, or enter the source manually. You can also paste an existing reference list into the citation checker. The tool keeps matched and unmatched entries visibly separate so an uncertain source is not presented as verified.

Metadata sources

• DOI lookups use DOI content negotiation, which can return records registered with Crossref and other DOI registration agencies.
• PubMed ID lookups use the NCBI PubMed citation API.
• ISBN lookups use Open Library.
• URL lookups read public page metadata when it is available.

An identifier is sent to the relevant lookup service when you request a lookup. The resulting citation list is stored in your browser's local storage; no account is required. See the Privacy Policy for details.

Pasted-reference matching

The citation checker accepts up to 20 references in one batch. It first looks for a DOI, PubMed ID, ISBN, or URL in each entry. An identifier is treated as an exact lookup path, although the returned record must still be compared with the source.

When a reference has no stable identifier, NeatCite sends the text to Crossref's bibliographic search. A result is accepted only when the candidate title has strong token coverage and an author surname or publication year also agrees, unless the title match is nearly complete. The result is labeled as a Crossref match and still requires your confirmation.

If neither path produces a reliable match, NeatCite does not reconstruct missing metadata from punctuation. It preserves the entry and applies only safe spacing cleanup. Add a stable identifier or enter the source fields manually before treating that citation as formatted.

Formatting engine and style rules

Formatting runs in the browser with the open-source citeproc-js processor. NeatCite bundles Citation Style Language (CSL) files for each style shown in the citation-style directory. These rule files define ordering, punctuation, name formatting, dates, journal details, and in-text citation form.

Where a style has a named publisher or professional body, its generator page links to primary guidance and the corresponding rule file. Bundled CSL files retain their original attribution and license metadata.

The embedded upstream update timestamps in the currently bundled files are 2026-02-06 for AMA, 2026-01-07 for NLM, 2022-01-13 for ASM, and 2024-01-24 for ACM. These dates identify the rule-file revisions packaged by NeatCite; they do not replace current course, journal, publisher, or venue instructions. The review date on a generator page records when NeatCite last compared its explanation and examples with the cited guidance.

Verification

Before release, automated checks confirm that every registered style has a readable CSL file and can render both a bibliography entry and an in-text citation from the same structured journal-article fixture. The checker parser is also tested against line-separated, blank-line-separated, and numbered reference lists, along with identifier extraction and conservative source-matching cases. Targeted pages include worked examples so differences between styles are visible.

Limits and your final check

Citation software cannot determine whether imported metadata is complete or whether an instructor, department, journal, or publisher has adopted a local variation. Web pages often expose incomplete authors, dates, or publication titles. A strong bibliographic match can still point to the wrong version of a work. Always compare the source match and fields with the original, then follow the instructions supplied for your assignment or publication.

If you find a reproducible formatting problem, email contact@neatcite.com with the citation style, source identifier, current output, and expected output.