ResearchRabbit is a browser-based literature review tool that starts from papers you already trust and recommends connected research by following citations, references and author links. In the company's words, ResearchRabbit analyzes papers you love and finds the research you’ve been missing.
It is a discovery and citation mapping engine rather than a writing assistant. The company says ResearchRabbit does not use large language models to power its core recommendation algorithms.
Its current terms of service are issued by Litmap Limited, which does business as Litmaps and is registered in New Zealand. The site footer carries the notice “Copyright © Litmap Limited. 2016-2026. All rights reserved.” When the redesigned version was announced, the company credited the change to its partnership with Litmaps, saying it would bring millions of new articles and smarter searches.
The company says ResearchRabbit is trusted by 1,000,000+ researchers worldwide. It has a free tier and a paid tier called RR+ (ResearchRabbit+).
Using one or more input articles, ResearchRabbit searches the local network of citations, which covers all the citations and references of those starting articles.
Searches grow in steps, because ResearchRabbit uses the articles you select to learn and improve the quality of its recommendations.
Connections are mostly citation links, shared references and shared authors, but ResearchRabbit sometimes infers a connection from the semantic similarity of titles and abstracts. Basic search settings decide where candidates come from: all types of connection to the seeds, only the references of the seeds, or only their citations. The same logic applies when the search is for authors instead of papers.
Results appear as a ranked list and on an interactive citation map. By default the best 20 articles are shown on the map and on the first results page, and you can change that to 10, 20 or 40 articles. Beyond single maps, ResearchRabbit lets you place papers on a timeline, trace how ideas rise and identify key authors in a field.
Keyword search accepts the boolean operators AND, OR and NOT, plus quotation marks for exact-phrase matches. The company says ResearchRabbit's keyword search uses the Google Scholar search engine, so its results should match what you would see in Google Scholar.
Shared collections had been missing since the new ResearchRabbit launched in October 2025 and were later brought back as part of the free tier.
A Zotero Importer brings papers saved in linked Zotero collections into your ResearchRabbit library. Selected papers can be exported as BibTeX, RIS or CSV files, including papers drawn from several collections at once.
The RR+ plan raises the seed limit to 300 articles and adds advanced search controls, multiple projects, Signals alerts and faster support responses. The advanced filters cover keywords, publication date, journal quality through quartiles and journal h-index, open-access PDFs, and retraction status. ResearchRabbit has partnered with Signals, a service that looks for high-risk papers based on their citation behavior and citation network, and its warnings appear inside RR+ in your collections and new searches.
The database covers 310+ million articles across disciplines, drawn mainly from Semantic Scholar, OpenAlex and Crossref. It includes both open-access and closed-access articles. The company says it updates the database weekly and prioritizes the most recent, highest-cited versions of articles.
A ResearchRabbit search can start from keywords, a DOI, an existing collection or an imported file.
The company positions the tool as a complement: ResearchRabbit is designed to work with traditional library resources, not replace them. Library databases stay the place for structured keyword searching and full text; ResearchRabbit finds related work that keywords miss.
ResearchRabbit fits people who learn a field by following citation trails rather than by writing long keyword queries.
It is a weaker fit as the main engine of a systematic review. A university library evaluation judged the tool best suited for inspiration rather than systematic evidence searching. It is also not open to anyone under 18, and the terms restrict commercial use without the company's approval.
ResearchRabbit pricing has three levels, and the paid RR+ price depends on your country.
| Plan | Listed price | Seed articles per search | Main inclusions |
|---|---|---|---|
| Free | $0 | Up to 50 | Unlimited searches and collections, sharing, Zotero importer, BibTeX import and export |
| RR+ | $10 per month on the annual plan; $12.5 per month on the monthly plan | Up to 300 | Advanced search controls, multiple projects, Signals alerts, faster support responses |
| Institution | On request | Not stated | Everything in RR+, volume discounts, LibKey integration, user management, usage statistics, dedicated support |
The free tier is described as complete enough for a literature review, including unlimited search, unlimited collections, collaboration, library uploads and up to 50 seed articles per search. Signing up for it needs no credit card.
The plan card lists RR+ at $10 per month on the annual plan and $12.5 per month on the monthly plan. The help center gives the totals as $12.50 USD per month, or $120 USD per year, a saving of 20%. The pricing FAQ, by contrast, gives USD$10 per month as the default RR+ price for the U.S., UK, Canada and other countries. It describes RR+ as a premium subscription with monthly payments, with access continuing until the plan ends even after cancellation. The FAQ and the plan card therefore describe the $10 figure differently.
For 100+ countries, RR+ is discounted according to a country parity group. Eligible users see a discount code automatically when they upgrade in the app and apply it at checkout. Under the terms of service, all payments are in US dollars. Subscriptions renew automatically unless canceled. Cancellation takes effect at the end of the current paid term. The terms checked for this review contain no refund provision.
A university library evaluation was first published in September 2025, before the October 2025 relaunch. It advised using ResearchRabbit as a springboard for exploration, not as a comprehensive or authoritative search tool, citing limited source transparency and replicability. It also found that search outcomes could vary with the input method, title or DOI, which affects replicability.
Yes. The free tier has no time limit and allows up to 50 seed articles per search; RR+ is optional.
The terms of service name Litmap Limited, doing business as Litmaps, as the company that operates the ResearchRabbit website, and the redesigned version was announced as the result of a partnership with Litmaps.
Its core recommendations come from citation and authorship networks, not from a large language model. Separately, the privacy notice lists Databricks and OpenAI as providers for its AI Products.
Not by its own description. The company calls it a complement to library databases, and the library reviewers cited above advise using it alongside systematic searching.