Research & PolicyFree
NIH Specific Aims Review Before Submission
A mock study section critique of your own Specific Aims and Research Strategy, scored under NIH's simplified framework, with specific revisions to make.
River's NIH application review reads your own draft the way a study section will. Paste a Specific Aims page or a Research Strategy and get a mock critique organized by NIH's simplified review framework: importance of the research, rigor and feasibility, and whether expertise and resources look sufficient. Each weakness comes with the passage it refers to and a specific revision, ranked by how much it is likely to move the score. It does not write new sections for you.
That last line is deliberate. Since September 2025, NIH has said it will not consider applications substantially developed by AI, or containing sections substantially developed by AI, to be the original ideas of the applicant. A tool that drafts your narrative puts the application at risk. A reviewer's eye on your own words does not. So this one critiques, questions and suggests, and every sentence that goes into the application stays yours to write.
This is for investigators, postdocs and research development staff preparing R01, R21, R03 and similar research project grant applications, ideally a few weeks before the internal routing deadline, while there is still time to rework an aim. If an application has already been scored and declined, a rejected application score review works from the summary statement instead. Before upload, a submission pre-check catches the attachment and formatting problems that can stop an application after it is submitted. Studies with human participants also need an IRB protocol narrative written from the same aims.
How NIH reviewers will read your draft
For due dates from January 25, 2025, most research project grant applications are reviewed under NIH's simplified review framework. Factor 1, importance of the research, covers significance and innovation and is scored from 1 to 9. Factor 2, rigor and feasibility, covers the approach and is scored the same way. Factor 3, expertise and resources, is judged only as sufficient or not, and reviewers must explain any finding that it falls short.
NIH's notice on fairness and originality took effect from the September 25, 2025 receipt date. It says applications substantially developed by AI will not be considered the applicant's original ideas, and that AI use detected after an award may be referred to the Office of Research Integrity. The same notice limits each principal investigator to six new, renewal, resubmission or revision applications per calendar year, so each one is worth getting right before it goes in.
Take an illustrative Specific Aims page with three aims, where Aim 2 analyzes samples that exist only if Aim 1 succeeds. A reviewer will call that a dependency and mark down rigor and feasibility, however strong the idea. A useful critique names the sentence, explains the risk in a reviewer's terms, and suggests a fix, such as a second sample source for Aim 2. The investigator then rewrites the aim in their own words, and the page stays within its one page limit.
How it works
Paste your draft
Your Specific Aims page, Research Strategy, or both, plus the funding opportunity if you know it.
Get the mock review
A study section style critique by factor, with estimated scores and the passages behind them.
Revise in your words
Work through the ranked revisions yourself, keeping every sentence in the application your own.
Review again in chat
Paste the revised section and ask whether each weakness has actually been addressed in the text.
What you get
- A mock critique organized by the three factors of NIH's simplified review framework
- Strengths and weaknesses tied to the exact passage in your draft that prompts them
- Estimated factor scores with the reasons behind them, clearly labelled as a rehearsal
- Specific revisions ranked by how much each is likely to change a reviewer's view
- Checks for interdependent aims, unclear hypotheses, missing rigor elements and page length
- Your own words kept, with questions and suggestions instead of ghost-written replacement sections
Common questions
Will it write my Specific Aims page?
No. NIH has said applications substantially developed by AI will not be considered the applicant's original ideas, so River reviews your draft instead. It points to passages, asks the questions a reviewer would ask, and suggests what to change. A short example phrase may illustrate a point, but the writing stays yours.
Are the scores real?
No, they are a rehearsal. Real scores come from your study section after discussion and depend on the other applications in the round. The estimated factor scores show which factor is weakest and why, so you know where revision time will pay off. Treat them as a ranking of problems, not a prediction of your percentile.
What page limits apply?
For most activity codes the Specific Aims page is limited to one page. The Research Strategy is limited to 12 pages for an R01 and 6 pages for an R21 or R03, unless the funding opportunity says otherwise. The review flags a draft that runs over and suggests what to cut first, based on what reviewers need.
Does it check rigor and reproducibility?
Yes, as part of rigor and feasibility. It looks for a clear hypothesis, a design that can answer it, power or sample size reasoning, attention to relevant biological variables such as sex, plans for unexpected results, and a timeline that fits the aims. Missing elements are named with the paragraph where they belong.
Can it review a resubmission?
Yes. Paste the summary statement with your revised draft, and the review checks whether each criticism is answered in the text and in your one page introduction to the resubmission. Criticisms that remain unanswered are listed first, because reviewers of a resubmission look for exactly those before anything else.
Is my unpublished research kept private?
Your workspace is private to your account, and River does not use your inputs to train AI models. Check your institution's policy on AI tools and unpublished ideas before pasting. Separately, NIH prohibits its peer reviewers from using generative AI on applications they review, a rule about study section members rather than applicants.
Sources
Built by
Chandler Supple
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.
NIH Specific Aims Review Before Submission
Describe what you need, add any files you have, and your workspace opens with the work already underway.