A panel, not one opinion
Separate reviewers read for methodology, statistics, citations, logic, ethics, reproducibility, novelty and writing.
Find out what reviewers will say before they say it
Upload your manuscript, thesis, proposal or literature review. Specialist AI reviewers leave comments, a score and a short list of what to fix first.
- Word or PDF
- A fix-first list, not 100 comments
- Remembers your last version
49,000+research reports completed
9,700+documents peer reviewed
3,500+documents improved or reviewed
950+university email domains
Feedback that ends with a finished paper
Know what matters most
The 20 worst issues become your fix-first list. Clear it and your score reaches about 85. Everything else is optional.
It remembers your last version
Upload your revision, even under a new file name, and see what you fixed, what you partly fixed and what is new.
How it works
- 1
Choose the review type
Proposal, Thesis, Manuscript or Literature. Each one is reviewed against its own standards.
- 2
Upload your document
Word or PDF, from your computer or your Library. You can also paste the text.
- 3
Fix, upload, repeat
Work through the list, upload the new version, and watch the score move.
Your score moves when you fix things
In our own test on a sample chapter, fixing part of the list took the score from 18 to 54, and fresh reviewers independently scored that version 53.
- Fixed, partly fixed, still open and new, counted for you
- New comments on text you did not change stay optional
- A clear finish: when the list is done, you are done
Everything an editor would ask
See the readiness score, standards applied, a reporting checklist, required elements, an action plan, strengths and recommended improvements.
- ICMJE and reporting-guideline checks for manuscripts
- Defence preparation for theses
- Download the report as PDF or Word
Before you send it out
| What matters | CitedEvidence | Asking a general AI chat |
|---|---|---|
| Who reads it | Eight specialist reviewers plus an editor overview | One general reply |
| Long documents | Reads the whole file, with page references | Often limited by length |
| Your next version | Compares it with the last one and tracks each issue | Starts again from nothing |
| Output | Score, fix-first list and a report you can download | A chat reply |
An AI review helps you prepare. It does not replace peer review by human experts. Compared with general AI chat assistants as a category on 4 November 2026. Their features change, so check them yourself for your own work.
Built for the way you work
Before journal submission
Catch what an editor may desk-reject for: missing statements, weak methods, unsupported claims.
Before your supervisor reads it
Fix the obvious problems first, so their time goes on your ideas.
Before your defence
Get thesis-specific feedback and defence preparation, chapter by chapter.
Questions
Does this replace real peer review?
No. It helps you prepare. Scores come from AI reviewers, so separate runs of the same file can differ, and only human reviewers and your supervisor can make the final call.
Will it keep finding new problems for ever?
No. Your score moves on a fix-first list of the worst issues. Issues noticed in text you did not change are marked optional and do not lower your score, so the review has a finish line.
How does it know it is the same document?
It compares the text and title with your earlier reviews, so a renamed or edited file is still recognised. If it is unsure, the Reviewer asks you.
What can I review?
Proposals, theses, journal manuscripts and literature reviews, as Word or PDF, or pasted text.
Can it fix the problems too?
The Fix with Improve button sends your document to Document Improvement, which suggests changes you can accept or reject.