A manuscript is never quite done. You've revised the draft, addressed the gaps you already knew about, cleaned up the methods section. Then you read it one more time the week before submission and start wondering what you're missing that someone else will catch immediately.
That's the specific problem Paper Review is designed for: not a replacement for peer review, but a targeted scan that surfaces the issues a journal reviewer would flag before they do.
You upload your PDF and pick a domain profile. There are five: Machine Learning, Biomedicine, Psychology and Social Science, Chemistry and Materials, and General. The profile matters; each one loads attack vectors specific to that field's common failure modes.
Four passes run in sequence. First, a vision scan of every figure and table, checking axis labels, color contrast, and whether your main result is legible at print size. Then a citation check: references are cross-checked against CrossRef for existence, and the most load-bearing citations get fetched to verify they actually support what you claim. Third is the adversarial panel. Four personas debate the manuscript: a Methodology Critic, Statistical Skeptic, Data Integrity Officer, and Editor-in-Chief, each probing from a different angle. The fourth pass synthesizes everything into a prioritized checklist.
For a 20-page manuscript, this takes 4 to 8 minutes.
What you get back is not a score. It's a list of specific concerns, ranked A/B/C by severity, each one quoted against the passage in your manuscript. Some findings will be wrong. AI reviewers make mistakes. But each finding cites the source passage, so you can verify every item against your text in under a minute.
The checklist format matters more than it sounds. When you're two days from a deadline, a narrative assessment is the last thing you need. What you need is: here are four things to fix, here are three to consider, here are two we flagged that can probably be ignored. You work down the list and decide.
The manuscript and the review are deleted from our servers the moment you retrieve the result. There is no account to pull it from later, so save the Markdown locally before closing the tab.
Paper Review is not a replacement for peer review. It fills the gap between "I think this is ready" and "I've confirmed I haven't missed anything obvious." For $9, that's worth doing before you send.