About Clasr
A manuscript leaves traces before any reader formally evaluates it.
A claim’s position. Where the evidence stops short of it. Whether a method is shown, or only described. These patterns are already present in the text before peer review begins.
01 Non-decisional by design
02 Visibility before judgment
03 Human reading stays central
Clasr reads a manuscript systematically and returns what it finds as labeled, located signals with clear severity levels.
It doesn’t make the decision. The report shows what the manuscript is signaling; what you do with it remains yours.
Reading layer
A structured report,
not a decision.
Authors get a map of what becomes visible across the manuscript: not an overall score, not a decision, and not a recommendation.
Clasr surfaces the signals. What you do with them is up to you.
Clasr will not
Cross the line into judgment.
- Score manuscripts. No overall rating, acceptance probability, or quality grade.
- Write reviews. Clasr does not impersonate a reviewer or generate a reviewer-style verdict.
- Rewrite prose. Revision remains with the author. In Author Mode, each finding includes a short “Consider” note to help guide what to revisit.
- Retain uploaded manuscript files. Uploaded manuscript files are deleted after processing and are not used to train AI models. Reports remain available according to your history settings.
Clasr will
Show you what’s there.
- Read across eleven sections, from framing and methods through results, discussion, figures, limitations, and reproducibility.
- Calibrate to the target journal tier. Different tiers use different levels of scrutiny, with Auto available when no tier is selected.
- Adapt to the field. Clasr adjusts the reading to disciplinary context and activates the checks that apply to the manuscript.
- Show the basis of each finding, including where it appears, how it is classified, and its severity.
- Present the reading in three modes: Author, Reviewer, and Editor.
- Track whether central claims remain consistent from introduction through conclusion.
Academic AI landscape
Where Clasr sits among other tools.
Some tools find and summarize relevant papers for literature discovery and systematic review. Others draft prose or generate reviewer-style critiques. A separate category checks technical correctness, such as proof validity, equations, or formal mathematical reasoning.
Clasr does something else. It makes a manuscript’s editorial and structural patterns visible before submission, without producing a score or decision.
Built for
Narrow and deep. Built for one task done well.
For researchers preparing manuscripts. For advisors and supervisors. For institutions, journal teams, and research organizations working across manuscripts at scale.
Clasr currently reads English-language academic manuscripts and works best with complete drafts. Partial submissions can also be read when needed. Sections that are not provided are treated as unavailable rather than inferred.
Clasr is not a proof checker, plagiarism checker, peer-review simulator, or publication outcome predictor.
Clasr stops at the signal layer.
It shows you what becomes visible under structured reading and leaves the next decision to you.
