Draft finished
phdflow reviews your paper before the venue does
Upload the PDF and pick the venue. Six specialists return accept or reject in about four minutes, with every issue and the section it sits in. The verdict on your first paper is free.
No paper yet
Is the idea already taken?
It reads Semantic Scholar and arXiv, follows the future work other papers asked for, and draws the graph of prior art and open gaps. It ends GO, PIVOT or PASS.
Ready to send
Where should this paper go?
Paste a title and an abstract. It ranks the venues that publish this kind of work, with the reach and the article charge OpenAlex has on record. Free, no account.
Already submitted
Hear about the paper that undercuts you
Watch a topic and get an email the day a new paper lands in it, so you hear about the one that undercuts you before a reviewer does.
- PIVOTIllustration
Close to three papers, open on one side
Solid threads are work that already exists. Dashed ones are the future work other papers asked for and nobody has done.
30 credits, €1.58 in a Pro pack3 to 7 minutes a run
- AUC 0.91 on 297 ICLR 2025 papers 96% accuracy, 1 false accept in 149 rejects
Your first paper gets a free preview: the verdict, the score and its most serious problem, written out in full. The full report is 60 credits, which is €9. No account and no password: you give an address and the report is sent to it.
0.38Overall score, ICLR 2025Reject IllustrationSoundness41%Significance36%Statistical29%Novelty52%error 4. ExperimentsMain claim is not supported by the ablation in Section 4
Suggested fix: State the claim as conditional, or add the ablation that isolates the contribution.
60 credits a full report120 credits with a second, deliberating pass. Credits do not expire.
How accurate is it, measured The benchmark behind the numbers
Both publish this kind of work
IllustrationQuantum reach 89th charge $215npj Quantum Information reach 92nd charge $3,790From a real run: three percentiles apart in reach, seventeen times apart in price.
Free, no account neededWednesday, 06:12
Illustration2 new papers on "guideline-grounded clinical QA"
One of them reports the baseline your Section 5 says nobody has run.
Sent the day it lands, not the week you next remember to search.
100 credits, €5.27 in a Pro packOne topic, watched until you stop it
Pricing
Prepaid credits, priced per run
No subscription. Credits work on any mix of Judge and Explore and do not expire monthly.
A missing baseline is a weekend of compute before the deadline and a whole cycle after it. A review is €9.
Single review
60 credits
1 full paper review
- 2 topic explorations
- €9.00 per review
No pack, no leftovers. VAT included, invoice issued.
Starter
420 credits
7 full paper reviews
- 14 topic explorations
- €4.14 per review
VAT included. Invoice issued for every purchase.
Pro
1500 credits
25 full paper reviews
- 50 topic explorations
- €3.16 per review
24% less per review than Starter. VAT included, invoice for every purchase.
Lab
credits, priced on request
80 full paper reviews or more
- 160 topic explorations
- Shared invoicing
Email us and we will quote it, with an invoice for your institution.
If your first review is not worth what you paid for it, write to us within 14 days and we refund it. One per account, no form to fill in.
If a run fails, the credits go back to your account automatically, with no email to write.
Full pricing, invoices and VAT. Buying for an institution? Email us before you pay , a VAT number cannot be added later.
Questions
Frequently asked questions
Will using this get my paper desk-rejected?
No. Venue policies target two things: reviewers running confidential submissions through an LLM, and authors passing off generated text as their own writing. This is neither. You run it on your own manuscript, before submission, for feedback. Disclosure rules are your venue's and they change every cycle, so read them: we are not going to claim blanket compliance on your behalf.
Where does my unpublished manuscript go?
Your PDF is parsed by Grobid in our own container, not a third-party service. The extracted text goes to Google's Gemini API for the agent runs, and results are stored so you can reload the review. You can delete a review and its content from your account. Need retention or sub-processor terms? Ask, and we will answer in writing rather than in marketing copy.
Is this just ChatGPT with a wrapper?
No. Separately calibrated agents score different dimensions, the venue's published criteria are loaded from a database, and an orchestrator synthesizes rather than invents. The result is measurable: AUC 0.91 on 297 ICLR 2025 papers, with the confusion matrix on the architecture page. A single prompt does not get there.
How do I know AUC 0.91 is real?
Three ways to check. The dataset is 297 ICLR 2025 papers scraped from OpenReview, the same ground-truth source the Stanford Agentic Reviewer preprint used. The confusion matrix is on the architecture page, so accuracy, precision and recall are recomputable. And the sample review is open without an account: read what it actually says before you pay for one.
Will it just tell me what I want to hear?
No. On ICLR 2025 it produced 11 false rejects against 1 false accept, so on that set it errs harsh, and precision when it says accept is 99%. It is not harsh everywhere: on NeurIPS 2023 the errors run the other way. Either way a reject is a prompt to look harder at what it flagged.
What if the review turns out to be useless to me?
If your first review is not worth what you paid for it, write to us within 14 days and we refund it. One per account, no form to fill in. Tell us what was wrong with it while you are there. A run that did not earn its price is worth more to us as a reason than as nine euros.
Why would I pay when Consensus and ResearchRabbit are free?
Those are retrieval tools and good at retrieval. They do not tell you whether your paper reads as a reject. The comparison is not a free search box, it is finding out at the decision notification. At €9 for a single paper, or about four euro each buying a pack, one caught missing baseline pays it back.
The honest test
Run a paper whose fate you already know. The preview costs nothing, so if the verdict matches what happened you found that out without paying.
phdflow does not replace peer review. It rehearses it, in minutes rather than months.
No subscription. Finance Foresight AI Kamil Szczepanik, NIP 7252336619, Poland.