ResearchRabbit
Last reviewed
Jun 4, 2026
Sources
18 citations
Review status
Source-backed
Revision
v1 ยท 2,512 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 4, 2026
Sources
18 citations
Review status
Source-backed
Revision
v1 ยท 2,512 words
Add missing citations, update stale details, or suggest a clearer explanation.
ResearchRabbit is a free literature-discovery and citation-mapping tool that helps researchers explore academic publications through interactive visual networks rather than keyword search. Built by a small team of developers in Seattle and opened to the public in 2021, it lets users assemble "collections" of seed papers and then surfaces related work, references, citations, and the authors behind a field as navigable graphs. The product is widely described as the "Spotify for papers" because, like a music-streaming service, it learns from the papers a user collects and recommends more reading in the same vein. ResearchRabbit drew attention among academic librarians for pairing a feature-rich interface with a "free forever for researchers" pledge, an unusual stance in a market where comparable tools charge subscriptions. In May 2025 the New Zealand company Litmaps acquired ResearchRabbit and has since folded its larger data sources and search technology into the tool while continuing to operate it under the ResearchRabbit brand.
ResearchRabbit is a citation-based literature-mapping application that runs in the browser. A researcher begins by adding one or more papers, by title, DOI, PubMed ID, or by importing a Zotero collection or a RIS/BibTeX file, into a collection. The tool then analyzes the citation and co-authorship structure around those seed papers and offers three kinds of recommendations: Similar Work (topically related papers), Earlier Work (references the seeds cite), and Later Work (newer papers that cite the seeds). Results are shown not as a flat list but as a graph, with nodes for papers or authors and edges for citation and collaboration links, which users can expand outward to follow a line of inquiry. The University of Ottawa librarians Victoria Cole and Mish Boutet, reviewing the tool in the Journal of the Canadian Health Libraries Association in 2023, summarized it as "a scholarly publication discovery tool supported by artificial intelligence."
A key point that university guides stress is that ResearchRabbit is not a generative AI chatbot. Deakin University's library evaluation describes it as "not a generative AI tool but a citation-based discovery aid," powered by "AI-powered machine learning algorithms that map and recommend scholarly articles and authors based on citation, co-citation, and co-authorship networks." In other words, its intelligence comes from graph-based machine learning over scholarly metadata, not from a large language model generating text.
ResearchRabbit was created in 2021 by a small team of three developers based in Seattle, Washington. The application has been credited to Krishnan Chandra (a co-founder who served as the technical lead), Ben Slater, and Mike Ma, and university library guides describe the software as "developed and periodically updated by Krishnan Chandra, Ben Slater and Mike Ma." The founders framed the project around their own frustration with the literature-review process; the company's own materials say, "We're researchers, and we've been there. The long nights, the messy desktop, and that feeling of being completely overwhelmed."
The tool spent its early life in a closed beta accessible only by invitation. The academic-librarian and open-access commentator Aaron Tay, writing in August 2021, noted that "until recently, access to it was limited via an invite system, but this changed last week," marking the point at which ResearchRabbit opened registration more broadly. By the end of 2021 it was available to anyone with an account, and it built a following among graduate students, librarians, and researchers who appreciated that a polished discovery tool was being offered at no cost.
From launch, ResearchRabbit positioned itself as free to use, and the Cole and Boutet review records that the platform stated it would remain "free forever for researchers." That commitment, combined with a deep feature set, made it a frequent recommendation in university research guides, several of which singled it out precisely because it did not gate features behind a premium tier. The original company was small (library evaluations and company databases describe it as roughly three employees, headquartered in Seattle) and kept a low public profile, releasing little formal information about its finances.
On 8 May 2025, the Wellington, New Zealand startup Litmaps announced it had acquired ResearchRabbit and completed the first close of a USD $1 million funding round to accelerate AI-driven research discovery. The raise was led by the UK-based group Scholarly Angels, represented by Andrew Preston, the founder of the researcher-recognition service Publons (which was acquired by Clarivate). Litmaps, founded in 2016 and led by co-founder and chief executive Axton Pitt, said the combination would push its user base past two million researchers across institutions such as Harvard, Stanford, and the University of Cambridge, and reported that its annual recurring revenue had roughly doubled to about USD $1 million ahead of the round. Pitt argued that "most tools for navigating academic literature are outdated, clunky or simply not built for the way modern science works," while Preston said "Litmaps is at the forefront of a number of companies that are disrupting traditional discovery with a combination of AI and citation network analysis."
After the acquisition, Litmaps kept ResearchRabbit running as a distinct product and began integrating its own infrastructure, bringing "millions of new articles" and improved search into the tool. Several members of the leadership now listed on ResearchRabbit's site, including Pitt and head of product Digl Dixon, are the Litmaps team, reflecting the merged organization.
In October 2025 ResearchRabbit rolled out a major release built on the Litmaps partnership. The update expanded the underlying database, added advanced filtering and reworked search workflows, introduced integrated note-taking, and refreshed the interface. The most significant change for the product's identity was the introduction of an optional paid tier alongside the free version. ResearchRabbit's announcement said the free version remains available with existing users' collections and features carried over unchanged, while a new Premium tier offers "more advanced searches, deeper connections, and even smoother workflows." The company emphasized "country-based pricing," charging based on local economies so the paid product stays accessible globally, and reiterated a commitment to remaining the free option for researchers worldwide. Reviewing the new version, Aaron Tay noted that the revamp brought ResearchRabbit and Litmaps much closer in capability than they had been before, given their now-shared technology.
ResearchRabbit is organized around collections and graphs. The typical workflow is:
Because the recommendations are iterative, the tool is generally considered more powerful once a collection contains several papers rather than just one. Reviewers have observed that this matches how experienced researchers actually conduct a literature review, expanding outward from a known core, and contrasts with single-shot visualization tools.
The defining feature is the interactive graph. ResearchRabbit renders citation and co-authorship relationships as visual maps with author names and publication years, letting researchers see clusters and connections that a ranked list would obscure. Beyond papers, it maps the people in a field: users can explore an author's body of work, view collaboration networks, and discover other influential researchers nearby. Cole and Boutet credited these maps with revealing "publication and author connections users might otherwise miss," and noted that the Similar Work feature can surface related publications that traditional keyword search does not return.
Papers are gathered into named collections, to which users can add private notes and annotations. Collections can be shared with collaborators in read-only or editable modes, supporting group literature reviews. This collaboration layer, along with the ability to keep papers, notes, and thoughts together, is part of how the product positions itself as a research workspace rather than just a search box.
Once a collection exists, ResearchRabbit can monitor it and send periodic email digests when new, relevant papers appear, described by early reviewers as weekly updates "as the algorithm finds new papers that might be of interest." The company characterizes these alerts as personalized and deliberately non-spammy, a way to stay current in a niche without manual re-searching.
ResearchRabbit offers a two-way integration with the open-source reference manager Zotero, which university libraries highlight as one of its most useful capabilities. Users can import an existing Zotero collection into ResearchRabbit to seed discovery, or build a collection in ResearchRabbit and sync it back to Zotero, with a "Resync with Zotero" option to keep the two in step. Notes carried in Zotero come across on import, and notes made in ResearchRabbit can be synced back. The integration is oriented around a user's personal Zotero library; support for Zotero group libraries has historically been more limited.
ResearchRabbit does not maintain its own crawl of the literature; it draws on large open scholarly metadata sources. Current university evaluations identify its primary sources as OpenAlex, Semantic Scholar, and PubMed, with PubMed powering search in the medical and life sciences and Semantic Scholar covering other disciplines. The company has claimed its combined corpus runs to hundreds of millions of articles, described in the Cole and Boutet review as "second in size only to Google Scholar," and the current site advertises access to "over 270 million academic papers."
The data foundation has shifted over time. When the tool launched, Aaron Tay reported that it "draws on the data made available in Microsoft Academic Graph (MAG)," supplemented by Lens.org and PubMed for keyword search. Microsoft Academic Graph was discontinued at the end of 2021, which created a coverage gap for newer publications until the tool migrated to OpenAlex and Semantic Scholar, sources that succeeded MAG and are kept current. The Litmaps acquisition further enlarged the database by adding Litmaps' own article coverage.
A recurring caveat in library assessments is transparency. Deakin University's evaluation notes that the provenance of "additional sources" beyond OpenAlex and Semantic Scholar "is unclear," that the Similar Work algorithm is proprietary and effectively a black box, and that the tool does not explicitly filter retracted papers, predatory journals, or misinformation. Author-name disambiguation can also occasionally produce duplicate nodes for the same researcher.
ResearchRabbit launched as, and for years remained, entirely free, requiring only a free account to use. Cole and Boutet recorded that "the tool is completely free." That changed with the October 2025 release, which introduced an optional Premium tier while preserving a free version. The company applies country-based pricing to the paid product so that cost scales with local economies, and continues to market ResearchRabbit as the free entry point for literature discovery. Exact prices vary by region and over time and are published on the official site.
ResearchRabbit has been received positively in academic and library circles, where it is frequently included in research guides and comparisons of literature-mapping tools at institutions worldwide. Reviewers praise the speed and relevance of its recommendations from a set of seed papers, the richness of its author networks, and, before 2025, the fact that all of this was free. Aaron Tay, comparing it with Connected Papers, found that Connected Papers "makes better recommendations than ResearchRabbit based on one paper" as a single-shot tool, while ResearchRabbit is built for iterative, multi-paper exploration. The Cole and Boutet review flagged usability limitations in earlier versions, including that exploration followed a single linear path with no easy way to branch and save alternative search routes or mark already-seen papers, some of which later updates and the Litmaps integration aimed to address.
ResearchRabbit sits among a cluster of citation-based discovery tools that emerged around 2020 to 2021. They share the idea of starting from seed papers and using citation links and recommendation algorithms to map a field, but differ in interface, number of seeds, and depth of organization.
| Tool | Approach | Seeds | Notable traits |
|---|---|---|---|
| ResearchRabbit | Iterative citation and author mapping with collections | Many | Free core, author networks, Zotero two-way sync, alerts; acquired by Litmaps in 2025 |
| Litmaps | Citation maps plus modeling of a researcher's existing knowledge | Many | More customization and sharing; combines citation analysis with AI; now owns ResearchRabbit |
| Connected Papers | Single-shot similarity graph around an origin paper | One (can add a second origin) | Fastest and simplest for a one-paper overview |
| Inciteful | Iterative network analysis across multiple seeds | Many | Highly customizable graph queries; no separate maps or collections |
Connected Papers is generally regarded as the quickest way to get a single-paper overview, Litmaps and ResearchRabbit offer richer multi-map and collection management, and Inciteful emphasizes flexible network analysis without persistent collections. Because Litmaps acquired ResearchRabbit, the two have grown noticeably more similar, sharing data and technology while keeping distinct front ends.