Free Online Citation Checker

Whether the source exists at all

The core check: does a scholarly record appear behind the citation. In our frozen 369-reference benchmark, no fabricated citation was classified as a source record found.

Hallucination risk

Flags citations that look plausible but cannot be confirmed as a real source. Measured on the same benchmark: 88.7% of flags were correct, with a 2.4% false-positive rate.

DOI resolution

Checks whether the DOI resolves to a real record. A DOI that resolves to a different work than the one you pasted is harder to catch and often passes.

Match evidence and coverage

Looks for records in indexes including CrossRef, OpenAlex, PubMed, Semantic Scholar, and Google Books. Grey literature — theses, institutional reports, non-indexed regional journals — can be real even when no match appears here.

What it does NOT reliably catch yet

Field-level errors in a source that genuinely exists. A wrong journal name, volume, page range, or year often still returns as matched. Verify those by hand before you submit.

Example input

Harris, C. R., Millman, K. J., van der Walt, S. J., et al. (2020). Array programming with NumPy. Nature, 585, 357-362. https://doi.org/10.1038/s41586-020-2649-2

Checker output

The DOI resolves and the title, author set, journal, year, and pages line up with the source record.

Example input

Harris, C. R., Millman, K. J., van der Walt, S. J., et al. (2019). Array programming with NumPy. Nature, 585, 357-362. https://doi.org/10.1038/s41586-020-2649-2

Checker output

When the checker catches it, a wrong year downgrades the status like this. Field-level errors are the hardest case, so treat a clean result on an existing source as "not disproven" rather than "confirmed correct".

Example input

Smith, J. (2024). Neural pathways in classroom memory. Journal of Applied Learning, 18(2), 44-61. https://doi.org/10.5555/fake.2024.118

Checker output

The DOI does not resolve and no matching title-author record appears in the checked databases.