Catch errors before they're published.One report your team can act on.

Figures, tables, statistics, and references checked, with the evidence behind every flag. For R&D, research-integrity, and publishing teams.

Sign up free

For publishers

www.reviewerzero.ai/app/review/8f3a91c2
SummaryAuthorsFiguresStatsRefsAI25 pages
Issue summary66 checks
6Review
11Suggest
49Clear
  • Author verification
  • Figure integrity
  • Statistical validation
  • Reference validation
  • Replicability
  • Plagiarism

Used by publishers and research institutions.

Featured in
Science
Nature
The Scientist
Partners & integrations
STM Integrity Hub
ScholarOne Manuscripts
Karger Publishers
CTI Meeting Technology
Lifecycle Journal
Recognition
SSP EPIC Awards, Silver, Integrity Tools 2026
Lab Venture Challenge, Winner 2023

Checked against the published record.

Each check compares your work with reference data we collect and maintain.

  • 35M+

    Published figures

    Used to find the same image in other papers and on the open web.

    Used by Figure Integrity

  • 17M+

    Published tables

    Finds table data that also appears in other papers.

    Used by Table Duplication

  • 1M+

    Peer reviews

    Finds review text shared with earlier peer reviews.

    Used by Review Quality

  • 240M+

    Scholarly records

    Confirms that each reference exists and matches what is cited.

    Used by Reference Validation

  • 15,000+

    Journals profiled

    Bibliometric, editorial-board, author, and institution data behind each journal's risk score.

    Used by Journal Monitor

  • 23,000+

    Authorship-for-sale offers

    Offers to buy a place on a paper, taken from paper-mill ads and matched against the literature.

    Used by Paper Mill Ad Watch

Twelve dimensions. 65+ checks.

From author identity to peer-review quality, each flag is traced to its source and explained.

  • Author Verification

    5 checks

    Confirms identities, affiliations, and publication histories

  • Figure Integrity

    6 checks

    Duplicated, altered, or AI-generated images within a paper, with cross-paper matching available

  • Panel Analysis

    3 checks

    Per-panel integrity and accessibility

  • Statistical Validation

    6 checks

    Recomputes reported test statistics and p-values for consistency

  • Reference Validation

    15 checks

    Hallucinated references, out-of-context citations, retraction status, and broken links

  • Citation Worthiness

    2 checks

    Claims in the text that lack adequate citation support

  • AI Text Detection

    3 checks

    AI-generated text, at the paragraph level

  • Plagiarism Detection

    3 checks

    Text that overlaps with other sources, traced to the source

  • Replicability Assessment

    6 checks

    Estimates the probability a study's results would replicate

  • AI Peer Review

    8 checks

    A structured, journal-quality review in minutes

  • Submission Compliance

    4 checks

    Checks against journal-specific submission requirements

  • Review Quality

    5 checks

    Depth of peer reviews, AI-written text, and review-mill patterns

Before it leaves the lab. Before you build on it.

The same checks, at four points in R&D work.

  • Your manuscripts

    Before submission

    • Recomputes reported statistics and flags results that do not match
    • Checks figures and tables for duplicated or altered content
    • Flags references that are retracted, can't be found, or don't match their source
    • Checks the target journal's submission requirements
  • The evidence behind it

    Before a deal

    • Checks the papers behind an asset or a partner's claims
    • Finds figures that also appear in other papers or online
    • Flags retracted work and statistical errors
    • Verifies authors and their publication histories
  • Partner and CRO data

    When results arrive

    • Screens blots, micrographs, and other images for duplicated or altered regions
    • Finds duplicated values across numeric tables
    • Recomputes the statistics in the report
  • Independent reviewers

    When you need experts

    • Finds experts on a manuscript's topic, ranked by fit
    • Flags co-authorship and other conflicts of interest
    • Confirms identities against publication records
www.reviewerzero.ai/app/review/8f3a91c2?tab=figures

Figure 3 · 6 panels

Web presence
1122
Filter tools

Findings

Copies2
  • Copy #1
  • Copy #2
Web reuse1
AI panels0
Accessibility0

Figures checked panel by panel.

ReviewerZero finds duplicated and altered image regions within a paper and shows exactly which regions match. Matching against the published literature and the open web is available too.

  • Image duplication

    Detects duplicated regions within a figure, including rotated, scaled, flipped, or skewed copies.

  • Image forensics

    Forensic filters (levels, thresholds, edge detection) reveal hidden changes in blots and microscopy.

  • AI-generated images

    Flags panels whose C2PA content credentials mark them as AI-generated.

  • Web & prior-publication matches

    Reverse-image search finds figures that already appear elsewhere on the web or in earlier papers.

Every reference, checked.

Fifteen checks on every reference: whether it exists, whether it is relevant to the paper, and whether it has been retracted. Hallucinated references, a common error in AI-assisted drafts, are flagged with the in-text context.

  • Hallucination detection

    Flags references that can't be found or don't match a published record.

  • Out-of-context citation detection

    Flags references that appear irrelevant to the manuscript's topic, methods, or claims.

  • Retraction status

    Flags references to retracted work.

  • Broken & restricted links

    Checks that every cited link resolves and is accessible.

www.reviewerzero.ai/app/review/8f3a91c2?tab=references
All58Hallucinated1Out of context3Retracted1Broken URLs0Restricted0

Showing 58 of 58 references

  • Hallucinated reference
    Citation context
  • Out-of-context citation
    Citation context
  • Retracted
    Citation context

Reported statistics, recomputed.

ReviewerZero recomputes reported tests from the numbers in the paper, shows the math, and flags results that don't match.

Needs reviewReported p 0.09 → computed 0.013 · F-test
www.reviewerzero.ai/app/review/8f3a91c2?tab=statistics
Statistical validation: reported vs. recomputed p-values with explanations
www.reviewerzero.ai/app/journal-monitor

Publisher portfolio

8 journals monitored

38% portfolio risk
62.8k
Articles
1,533
Flagged
24.4/1k
Problem rate
4 / 8
At risk
JournalFlaggedRisk
  • Cellular Imaging Reports
    Cell biology
    412High
  • Pharmacological Intervention
    Pharmacology
    337High
  • Oncology Advances
    Oncology
    253Elevated
  • Materials Science Frontiers
    Materials
    120Moderate
  • Geosciences Review
    Geoscience
    57Low
Contributing factors
  • Image integrity+14%
  • Author network+11%
  • Text & stats+8%

Know which journals carry integrity risk.

Journal Monitor scores integrity risk for each journal, so editorial teams can focus on the titles that need attention and R&D teams can vet the journals they publish in and cite. The feature is in early access.

  • Per-journal integrity risk scoring
  • Flagged-article and problem rates across a portfolio
  • Editorial-board, author, and institution signals behind each score
  • Early signals before issues become public

A structured peer-review report, in minutes.

A summary and prioritized issues, written for editors, reviewers, and internal review teams.

  • Summary and prioritized issues (critical, suggestion, low)
  • Replicability prediction (optional)
  • Journal-fit recommendations (optional)
  • Submission-guideline checks (with a target venue)
www.reviewerzero.ai/app/review/8f3a91c2?tab=peer-review
Major revision
3 critical7 suggestions51 passed

Summary

Issues

AllNeeds reviewSuggestionLow

How we handle unpublished work.

What happens to your manuscripts, figures, and data after you upload them.

Security and data protection
  • Not used to train AI models

    Your content is not used to train AI models, by us or by the service providers we use to process it.

  • Deleted after 7 days by default

    Uploads in the app are deleted 7 days after upload unless you turn deletion off. Organization admins can require it for every analysis.

  • Encrypted in transit and at rest

    TLS 1.2 or higher in transit and AES‑256 at rest.

  • You decide who sees results

    Reports stay in your workspace until you share them with your team, specific people, or a link.

  • You choose what runs

    Checks that search the web or scholarly databases can be turned off for any analysis, or for a whole organization.

What integrity teams say.

From publication-ethics and editorial teams.

  • “Manual verification of statistical findings is time-consuming. ReviewerZero's statistics tool is flexible, saving time and explaining its output in an easily digestible manner for research-integrity teams and editors.”
    Dr. Gráinne McNamara
    Publication Ethics Manager, Karger Publishers
    Karger Publishers
  • “We are thrilled to partner with ReviewerZero AI. ReviewerZero provides key insights on image integrity, statistical accuracy and citation analysis, aligning with our values of openness and transparency.”
    Dr. Eileen Clancy
    Managing Editor, Lifecycle Journal (Center for Open Science)
    Lifecycle Journal
  • “ReviewerZero helped us catch problems we would otherwise have missed, in papers that ultimately went on to be rejected. Our editorial offices particularly value the reference checks, which surface hallucinated references and citations that do not support the statements they are attached to. It helps keep our publishers' portfolios clean, and that is exactly what they want.”
    Dr. Lisa Clancy
    Research Ethics and Integrity Manager, Compuscript
    Compuscript

See it on your own manuscripts.

In a 30-minute demo, we run a few of your manuscripts and walk you through every flag. Papers you have already published work too.

Sign up free

For publishers