AI tool that summarizes long research papers into key points.
Research
Scholarcy is the AI paper summarizer built for the specific reality of research work — piles of PDFs, a deadline for a lit review, a supervisor who wants a synthesis by Friday, and a stack of dense methods papers you have to somehow triage before you can decide which ones to actually read. Founded in 2018 in the UK by Emma Warren-Jones and Phil Gooch, Scholarcy sits at the intersection of "summarizer" and "structured extractor" — it does not just spit out a paragraph summary, it produces a structured summary card with the paper's contribution, methods, key findings, limitations, and (crucially) auto-generated flashcards you can review or export. It is the tool grad students use when they have twenty papers to get through and three days to do it.
What Scholarcy actually does, cleanly stated, is turn any research paper into a structured summary card in about 60 seconds. You upload a PDF (or paste a URL, or install the browser extension and click "summarize" on an open paper), and Scholarcy extracts the key claims, the methods used, the sample size and dataset, the results, the limitations, and the referenced works — and packages them into a card with sections you can scan in two minutes instead of two hours. The summary is anchored to the source, with direct excerpts pulled from the paper for the important passages so you can verify what the tool is telling you.
The one-line positioning: Scholarcy is the paper triage tool for researchers who need to decide which papers are actually worth a full read. It is not the "chat with a PDF" tool that SciSpace is. It is not the systematic-review extraction tool that Elicit is. It is the fast-scan tool that turns a folder of 40 PDFs into 40 summary cards you can sort through in an afternoon — deciding which four papers deserve a careful read, which twelve you can cite from the summary, and which twenty-four you can safely skip.
Scholarcy has expanded meaningfully since launch. What started as a PDF-summary browser extension is now a full research assistant with library management, structured extraction across multiple papers, comparison matrices, and integration with reference managers (Zotero, Mendeley, EndNote). Pricing has consolidated around a freemium model with a $9.99/month Personal Library tier that most working researchers land on.
Scholarcy's product surface is deliberately extraction-first. The features that earn the subscription are the summary card format, the flashcard export, and the personal library for managing dozens or hundreds of summarized papers.
Summary Flashcards. The signature feature. Every paper Scholarcy processes becomes a structured card with sections: contribution, background, methods, dataset, results, key findings, limitations, and referenced works. Each section is short (one or two sentences), extractive (pulled directly from the paper), and scannable in about two minutes. This is the workflow shortcut that grad students actually pay for.
Highlighted key phrases. Scholarcy identifies the sentences in the paper that carry the most information — the claim sentences, the results sentences, the limitation sentences — and surfaces them separately. Useful for the "what does this paper actually say" first pass.
Referenced works extraction. Scholarcy pulls the reference list into a structured format and identifies which references are the most heavily cited within the paper (the "key references"). Useful for backward-citation tracking when building a lit review.
PDF and URL import. Upload any PDF, paste a DOI or arxiv URL, or use the browser extension to summarize any paper you have open. Scholarcy handles most journal PDFs, preprints, and open-access papers cleanly. Occasional issues with scanned older papers where OCR quality is poor.
Personal Library. A workspace where all your summarized papers live. Tag them, sort them, search across summaries, and export the library in bulk. This is the feature that turns Scholarcy from a "summarize one paper" tool into a real research workflow.
Comparison mode. Select multiple papers and Scholarcy produces a side-by-side comparison matrix — same summary sections across papers, so you can scan sample sizes, methods, and findings across a set of studies. Not as deep as Elicit's structured extraction, but faster for a small comparison set (5-15 papers).
Flashcard export. Export summary cards to Anki, Quizlet, RemNote, or plain CSV. Genuinely useful for students preparing for exams or researchers building a personal knowledge system.
Reference manager integration. Direct export to Zotero, Mendeley, EndNote, and RefWorks. Not just the reference metadata — the summary card content flows into your notes field, which means your reference manager gains actual content per entry instead of just citations.
Browser extension. Chrome, Firefox, and Edge extensions that summarize any paper you have open on ArXiv, PubMed, Semantic Scholar, or any publisher site. This is the workflow shortcut for the "I'm already on the paper, just summarize this" moment.
Export formats. DOCX, XLSX, RIS, BibTeX, JSON. Data portability is real — every summary you generate can leave the tool cleanly.
What Scholarcy deliberately does not do. No chat interface — you cannot ask questions of a paper the way SciSpace lets you. No paper search or discovery — you bring the papers to Scholarcy, not the other way around. No writing assistance for drafting a lit review. The narrowness is intentional.
Scholarcy is freemium with a limited free tier and a Personal Library subscription that most working researchers land on.
| Plan | Monthly | Annual (per month) | What you get |
|---|---|---|---|
| Free | $0 | — | 3 summary cards per day, browser extension, basic export |
| Personal Library | $9.99 | $7.99 ($95.88/year) | Unlimited summaries, library, comparison, all exports |
| Institutional | Custom | Custom | Multi-user licenses, SSO, admin dashboards, campus access |
Free at $0 gives you 3 summary cards per day with the browser extension and basic export. It is a real usable tier for casual users — a professor summarizing one paper for a class, a journalist checking a study, a curious reader triaging one paper at a time. For anyone doing more than three summaries a day, the daily cap will bite immediately.
Personal Library at $9.99/month (or $7.99 effective on annual, about $95/year) is the working-researcher tier. Unlimited summaries, unlimited library size, comparison mode across multiple papers, all export formats, and reference-manager integration. This is the tier most grad students, postdocs, and early-career researchers land on within the first week of serious use. At $95/year, it is priced roughly in line with other academic AI tools — cheaper than Perplexity Pro, comparable to SciSpace Premium.
Institutional is quote-only for universities, hospital systems, and research institutions signing multi-seat contracts with SSO, campus IP-based access, and admin dashboards. Aimed at library sciences departments and research administration offices.
Scholarcy occasionally runs student and educator discounts through campus resellers — worth checking with your university library before committing to individual pricing.
The honest read: at $8/month annual, Scholarcy Personal Library is priced fairly for the triage workflow it enables. If you regularly have to read (or decide not to read) more than a handful of papers per week, the tool pays for itself in a single afternoon of saved reading time. If you read one paper a month and read it carefully, the free tier is fine.
Pros
Cons
Grad students doing lit reviews. The canonical use case. You have 30-60 papers to triage for a thesis chapter or a comprehensive exam, three weeks to do it, and no way to read all of them carefully. Scholarcy compresses the triage into an afternoon — summary card each, sort into "read carefully," "cite from summary," and "skip." This workflow alone justifies the subscription for any active grad student.
Early-career researchers keeping up with the literature. Postdocs and junior faculty who need to stay current across 20+ papers a week in their subfield. Scholarcy plus RSS feeds from ArXiv or PubMed becomes a weekly triage habit.
Students preparing for comprehensive exams or qualifying exams. The flashcard export to Anki turns Scholarcy into a study system, not just a summarizer. This is one of the strongest niche uses of the tool.
Science journalists and communicators. When a piece requires reading five recent papers on a topic in one day, Scholarcy makes it possible. Not deep-read quality, but adequate for background and quote-mining. Serious journalism still requires reading the papers you cite.
Product managers in regulated industries. Healthtech, biotech, and pharma PMs reading foundational papers to understand claims. Scholarcy plus SciSpace is a strong two-tool stack for the "read the actual papers behind the marketing" workflow.
Clinicians catching up on trials and reviews. Between patients, on a phone or tablet, with 10 minutes to spare. Scholarcy's summary cards fit the actual clinician reading budget better than a full paper.
Corporate librarians and research support staff. The person whose job is to feed evidence up to a leadership team. Scholarcy makes the "here are the three papers you should read and here are quick summaries of the other twelve" report tractable.
Bad fit. Anyone whose research is genuinely single-paper depth — you need SciSpace or a careful read, not a summary. Anyone doing systematic reviews at scale — Elicit is the specialist. Anyone whose papers are in humanities and non-STEM fields where Scholarcy's structured extraction is a poor fit for essay-style scholarship.
Scholarcy's real competition is other academic AI tools, each with a different angle on the same problem.
SciSpace — the chat-with-PDF competitor. SciSpace enables conversation with a paper; Scholarcy generates structured summary cards. Different phases of research. Many working researchers use both — Scholarcy for triage, SciSpace for the papers that made the cut and deserve a careful read. The two-tool stack is common for around $16/month combined annual.
Perplexity with Academic Focus — the "quick answer from the literature" alternative. Perplexity is faster for a specific question ("what does the evidence say about X"); Scholarcy is stronger for "give me the summary of this specific paper." Different jobs — one is search, one is summarization.
Elicit — the structured extraction competitor. Elicit is stronger for cross-paper systematic extraction (sample sizes, effect sizes, methods across 50 papers); Scholarcy is stronger for single-paper deep summarization. Elicit is more powerful for lit reviews at systematic-review depth; Scholarcy is faster for the working grad student.
Consensus — the evidence triage competitor. Consensus tells you what the evidence overall says on a question; Scholarcy summarizes individual papers. Complementary — use Consensus for triage of a question, Scholarcy for triage of the papers you find.
ChatGPT and Claude with PDF upload — the frontier-model alternative. Both can produce paper summaries with a good prompt. What they lack is the structured card format, the reference manager integration, and the library. Fine for a one-off summary; poor for a workflow that involves 30 papers over three weeks.
ExplainPaper — a smaller competitor focused on highlighting-and-explaining specific passages. Narrower than Scholarcy, cheaper, useful for the "I don't understand this equation" moment. Not a triage tool.
QuillBot Summarizer and general summarization tools — generic alternatives without academic tuning. Fine for casual summarization; miss the paper-structure awareness (methods, limitations, referenced works) that Scholarcy provides.
Notion AI or Obsidian with an AI plugin — the "I want to summarize in my existing notes tool" alternative. Fine if you already live in Notion or Obsidian and are okay with generic paragraph summaries. Weaker on the structured extraction that makes Scholarcy specifically useful.
Sign up free at scholarcy.com. No credit card required. Free tier gives 3 summary cards per day — enough to test the workflow on real papers.
Install the browser extension. Chrome, Firefox, or Edge. The extension is the workflow shortcut for the "I'm already on the paper" moment. Without it, you are pasting URLs and uploading files, which is slower.
Summarize three real papers from your reading list. Not toy examples. Pick three papers you were going to read carefully, run them through Scholarcy, and see if the summary cards match what you would extract yourself. This is the honest evaluation.
Try the flashcard export to Anki or Quizlet. If you are a student, this is the test that either sells the tool or exposes its limits. Export a paper's summary as flashcards, review them for a week, and see if the recall matches what you would want.
Set up the reference manager integration. Zotero, Mendeley, EndNote, or RefWorks. This is the feature that makes Scholarcy sticky — every paper you summarize automatically enriches your reference manager with actual content.
Try comparison mode on 3-5 papers on a related topic. This is the workflow that starts to feel like a lit review. Judge whether the side-by-side matrix accelerates your synthesis.
Upgrade to Personal Library if the free cap bites. At $8/month annual, the subscription is a fair call for anyone whose research includes triaging more than a handful of papers per week.
Export your library periodically. Every academic tool eventually raises prices or shuts down. Export to DOCX or CSV monthly if the library is important to your work.
Is Scholarcy better than SciSpace? Different jobs. Scholarcy is faster for triaging many papers via structured summary cards. SciSpace is deeper for reading a specific paper carefully via chat. Serious researchers use both — Scholarcy for the "which papers matter" pass, SciSpace for the "help me understand this one" pass. Complementary, not competitive.
How accurate are the summaries? For cleanly structured journal PDFs, the summary cards are accurate and extractive (pulled directly from the paper). For scanned or non-standard papers, quality degrades. Always verify specific numbers or claims before citing them — the summary is triage, not verification.
Which AI models does Scholarcy use? A mix of extractive models fine-tuned for scientific text plus frontier LLMs for the more generative sections (contribution, limitations). Model versions rotate as new releases land.
Does Scholarcy work with paywalled papers? Yes, if you upload the PDF yourself. Scholarcy does not provide access to paywalled papers. You bring the paper (via institutional access or open-access preprint), Scholarcy summarizes it.
Can I use Scholarcy on non-English papers? Primary support is English. Other major European languages (Spanish, French, German, Portuguese) work at reduced quality. Non-Latin-script languages are not well-supported.
Does Scholarcy train on my uploaded papers? Personal-tier data may be used for product improvement under the standard terms; institutional contracts typically opt out. Review the current privacy policy for specifics before uploading sensitive unpublished work.
Is Scholarcy good for humanities papers? Less so. The structured extraction (methods, sample size, results) is calibrated for empirical STEM research. Essay-style humanities scholarship does not fit the card format cleanly. Use with lower expectations for philosophy, history, and literary criticism.
Can I share my library with a colleague? Personal tier is single-user. Sharing requires an institutional license or manual export.
Scholarcy Personal Library at $8/month annual is the right pick for grad students, postdocs, and early-career researchers whose weekly reality includes triaging more papers than they can carefully read. The structured summary card format is genuinely differentiated — faster than a paragraph summary, more scannable than an abstract, more useful than a chat interface for the specific "which of these 40 papers deserves a full read" question. Combined with the flashcard export for students and the reference manager integration for working researchers, Scholarcy earns its subscription at a price that respects a grad-student stipend.
Where Scholarcy stops being the right answer: careful single-paper reading and comprehension (SciSpace is the specialist), systematic-review-depth structured extraction across many papers (Elicit is stronger), evidence-based question answering (Perplexity with academic focus, or Consensus, is faster), and humanities scholarship where the structured card format fits poorly. Scholarcy is a triage tool — honest about that scope, priced accordingly, and best used as one piece of an academic AI stack rather than the whole thing.
The honest recommendation: sign up free, summarize three real papers from your actual reading list this week, and judge whether the cards match how you would extract the same content yourself. If yes — and for most working researchers reading STEM literature, they do — Personal Library at $8/month annual is a fair subscription. Stack it with SciSpace at $8/month for the "deeply read the papers that made the cut" flow, and with Perplexity or Consensus for the "find the relevant papers in the first place" flow. The three-tool academic AI stack costs about $25/month annual combined and covers the full research reading workflow — from question to relevant papers to triage to deep read — in a way that ChatGPT alone does not match. Scholarcy is the triage layer, and it earns its place.