LitSourceLitSource

Primary tool: Free Citation Checker. Cross-check a reference list against database records and flag inconsistencies for human review.

Free AI Citation Checker — Check References and DOI Metadata

Check whether AI-generated or hallucinated references point to real papers that exist in scholarly records. Compare title, authors, publication year, journal, and DOI with traceable scholarly records. Coverage is strongest for biomedical literature; a match does not establish claim support.

Sign in for 10 credits per month. Verification costs 3 credits per batch of up to 10 references; the Free plan supports up to 10 references per run.

Paste Format Example
Garcia, M. (2022). Deep learning approaches to clinical trial recruitment. Nature Medicine, 28(4), 731-742.
Detected: 0 referencesEstimated cost: 3 credit(s)Sign up and get 10 credits
Checking account access and input limits…

What does a citation check tell you?

Separate publication existence, metadata agreement, and claim support. Record checking does not replace reading the paper. Coverage is strongest for biomedical literature and varies by source.

Metadata match
The checked fields agree. Read the source separately to assess claim support.
Metadata conflict
Open the source and correct the title, authors, year, or DOI before citing.
No record found
This is not proof of fabrication. Check spelling, identifiers, index coverage, and provider availability.

Worked example: a real DOI with the wrong year

The year below was deliberately changed to 2033 for this demonstration. Do not cite this altered entry.

Lincoff AM. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. New England Journal of Medicine. 2033. doi:10.1056/nejmoa2307563
Frozen-source regression output and original record
FieldSubmittedSource record
Year20332023
DOI10.1056/NEJMoa2307563
Regression outcomemetadata_mismatch: Candidate exists, but publication year conflicts.

Next: open the DOI source, confirm the year and remaining fields, then read the paper. This outcome does not establish support for any specific claim.

Public test method and case-by-case results

Run date: 2026-10-22. These 24 curated regression cases use fixed provider responses. This is not a live detection rate, an independent evaluation, or a representative accuracy benchmark. Some variants derive from only two real papers. Coverage gaps, retraction signals, and outages are simulated conditions, not claims about those papers.

Frozen-response checks (pass means expected behavior)
Case categoryCasesExpected behavior passed
Accurate references and DOI inputs66/6
References without a DOI22/2
Incorrect publication years22/2
Conflicting identifiers22/2
Invented titles with real DOIs22/2
Synthetic absent records22/2
Simulated coverage gaps22/2
Incomplete reference inputs22/2
Simulated retraction signals22/2
Simulated provider outages22/2

Download inputs, expectations, and fixed sources (JSON)Download case-by-case run results (JSON)

Free credits and your next check

After sign-in, the Free plan includes 10 credits per month. Verification costs 3 credits per batch of up to 10 references; free accounts can submit up to 10 references per run. See pricing for additional credits and paid account limits.

Capabilities

Move from a claim to traceable sources with less friction

Find supporting literature, inspect source details, and continue with the right research workflow.

01
AI biomedical reference screening
Review AI-generated clinical and medical citations before they enter your systematic review or medical manuscript.
02
PubMed-backed existence checks
Cross-check AI citations against PubMed and biomedical databases for real paper records matching title, authors, journal, and year.
03
Clinical metadata mismatch detection
Spot fabricated medical references where DOI or title points to wrong papers or non-existent clinical trials.

Use cases

How people use it

Use LitSource when you already know which claim, citation, or reference list you need to investigate.

1
Check AI-assisted medical drafts
Screen biomedical references added during AI-assisted clinical writing before submitting to medical journals.
2
Audit clinical bibliographies
Find suspicious medical citations in systematic reviews, meta-analyses, or clinical guidelines generated with AI.
3
Protect clinical submission quality
Reduce the chance that fabricated biomedical references reach medical journal reviewers or clinical guideline panels.

FAQ

Questions before you start

Review common questions before searching for sources or checking citations.

The Free plan includes 10 credits per month after sign-in. Verification costs 3 credits per batch of up to 10 references. Free accounts can submit up to 10 references per run; this is a credit allowance, not unlimited free checking.

Yes. AI can generate convincing but fabricated clinical trial references, PubMed-like citations, and medical journal entries that do not exist.

No. It gives you a faster first pass before you inspect important clinical sources yourself, especially for evidence-based medicine claims.

LitSource opens the verification workspace so you can paste biomedical references and review the results.

Related workflows

Continue with nearby research checks

Move between source discovery and citation verification workflows without returning to the home page.

Hallucinated citations checkerOpen workflowChatGPT fake references checkerOpen workflowFake citation checkerOpen workflow

Ready to continue

Continue with the right research tool

Start searching for sources or checking citations in LitSource.

Get startedReview capabilities
Free AI Citation Checker for Hallucinated Citations