Evidence HubType a topic or a question. Evidence Hub asks PubMed, Europe PMC, Semantic Scholar, ClinicalTrials.gov, OpenAlex and arXiv together, merges the duplicates, and puts the papers that answer you at the top, each with an AI summary and a data sheet. An academic search for students, researchers and clinicians, and a quick start for a systematic review or meta-analysis.
No credit card needed. Free to start: new accounts get points for AI, and you can search for studies free as often as you like.
749 studies (+ 628 less relevant) for dexlansoprazole in gastroesophageal reflux disease
Built for anyone who has to read the literature quickly and still trust what they read.
Type a drug, a topic or a full question, in medicine or any other field. It goes to every source at once, so you search for studies in one place instead of six websites. A study found by several of them becomes one card that keeps the longest abstract and lists each source that had it.
Each study is compared with what you typed, then the top results are scored precisely by AI. Weaker matches wait in a collapsed section, so your time goes to the papers that answer the question.
Every study is sorted by field, publication type, approach and study design, and the filters fit what comes back. For medicine and life sciences you also get evidence-based filters: design from meta-analysis to case report, evidence level, subjects (human, animal, in vitro, in silico) and safety or efficacy. Useful when you screen studies for a systematic review or meta-analysis.
749 studies (+ 628 less relevant) · sorted by Relevance
Every study gets a plain-language summary and a data sheet filled in from the abstract and, when there is one, the full text. The data sheet wording fits the field, so a physics paper and a clinical trial each show what matters for them.
Search for studies PDF-first: Evidence Hub looks for a legally open-access copy of each paper, then shows its figures and tables. For a paywalled paper you already have, upload the PDF yourself; you earn points when it passes the check.
Open a study and ask it questions. Answers come only from that paper's own text, with page citations that open the source PDF. Never a guess, never another study's data.
This paper is a systematic review and meta-analysis of existing studies, not an original trial that collected its own adverse event data.
Its Discussion p. 16 says the 60 and 90 mg doses were “well tolerated … without dose-dependent adverse events.”
Save any search and Evidence Hub looks again every week, then tells you about new studies in the bell and by email, newest first. Only recent work counts as new, so an old paper that only now turns up is not announced. Pause or stop an alert in one click, even from the email without signing in.
A Continue card brings back your last search, with its filters and your place in the list, and your recent searches are one click away. Spelling is corrected as you type, and a Compact view shows results as a grid with shorter summaries when you want to scan fast.
Save studies into folders you create and drag them around. Export the list to CSV or RIS for your reference manager when you start writing. Public study pages can also be found through Google and opened without an account.
Medical and life-science studies get an evidence level from 1, the strongest, to 5. It is worked out from the study's design and who it was done on, so you can filter to meta-analyses and randomized trials first and read down from there.
Levels are assigned automatically from the abstract and can be wrong. Check the paper before you rely on one.
Find research articles in any field and read a plain-language summary before you decide what to open. Export what you keep to RIS or CSV for your reference manager.
Filter to human studies at evidence level 1 or 2, see safety and efficacy at a glance, and check the registry record of a trial without leaving the page.
Search the major databases at once, filter by study design, and export the list as a first step. It does not replace a protocol-driven systematic review, so keep your own screening record.
Put your topic or question in one search bar. You don't pick a database first, and you can untick any source in the filters.
Results are sorted by how well they match, with an AI summary on every card and the weakest matches folded away.
Open a study for its data sheet, chat with the paper, or save it and keep searching.
Searching is free. AI work (summaries, data sheets, Ask AI, topic overviews) is paid for in points: every new account starts with a welcome balance, and you earn more by uploading a PDF that passes the check. We're still finalizing plans and will announce changes ahead of time.
$0
Coming later: Pro, with all AI features free of points, priority full-text retrieval and team folders. Team and Institution, with seats for a lab, admin and usage controls, and priority support.
Yes. Searching all six sources, filtering, saving studies and exporting are free, so you can search for studies free and online. AI features (summaries, Ask AI, data sheets, topic overviews) cost points; every new account starts with some, and uploading a verified PDF earns more. We'll announce pricing changes ahead of time.
Yes. It is a website for research for students as much as for professionals: one search bar finds research articles in medicine, biology, chemistry, physics, engineering, computing, social science and the humanities, with a plain-language summary on each so you can decide what to read first. Export to RIS or CSV for your reference manager.
It is a fast first step. Search all the major databases at once, filter by study design (meta-analysis, randomised trial, cohort and more) and evidence level, read the data sheets and export the list. It does not replace a protocol-driven systematic review: record your own screening and search methods, and check every paper.
Yes. Press Get alerts above any results list and Evidence Hub saves that search with its filters, checks it every week (or every day), and tells you about new studies in the bell and by email, with a link to the search sorted newest first. Free accounts can keep a few alerts; you can pause, change or stop one at any time.
When a legal open-access copy exists, Evidence Hub finds it and opens the PDF with page-cited answers. Search for studies PDF-first, and upload the PDF yourself when a paper is paywalled and you have access.
Every search queries PubMed, Europe PMC, Semantic Scholar, ClinicalTrials.gov, OpenAlex and arXiv by default. You can untick any source in the search filters.
No. The same search bar covers every field, and the filters (field, publication type, approach, study design) fit the results. When medicine or life-science studies are among them, evidence-based filters such as evidence level and subjects (human, animal, in vitro, in silico) appear too.
Each study's title and abstract are compared against your search with an embedding model, then the top results are scored precisely by AI. Results are sorted by that match, not by date or citation count, unless you change the sort order.
Summaries are grounded only in the study's own abstract and, when available, its verified full text — the model is told to say when the material doesn't answer a question rather than guess. Every summary is labelled with the provider that wrote it. Always check the original paper; this is not medical advice.
Not usually — Evidence Hub automatically looks for a legally open-access copy first. For paywalled papers you already have access to, you can upload a PDF yourself; uploads stay private to you.
Free copy: we download the open-access version when one exists — a PDF, or the article's full text from PubMed Central. To find it we also ask Unpaywall, which requires a contact email: yours is sent with that request. Paywalled? Upload the PDF you have access to. Every upload is checked first (the DOI printed in it, that it is the whole paper, and an AI read); one that fails is not used and you are told why. Uploads of paywalled papers stay private to you; an upload of an open-access paper is shared with everyone once it passes.
Your saved studies, folders and chat history are private to your account. A PDF you upload for a paywalled paper stays private to you; an upload of an open-access paper is shared with everyone once it passes the checks. If a PDF is flagged and you ask for a review, an admin can look at it.
Create a free account. No credit card needed.