Note: OpenAI leads the general AI space, but AI companies are developing deep research tools and experimenting with AI-powered academic searching in support of research. Perhaps you have faculty or students asking you to present these tools to classes.
"...Consensus offers an interesting approach to using AI to help researchers navigate scholarly literature... and [may be] one of the few that offers institutional subscriptions. Libraries interested in adding AI-driven tools to their collection may want to investigate further into whether Consensus would be a good fit for their users." — Faix, 2025
Consensus - http://consensus.app is an AI/LLM powered search engine designed to streamline academic research by providing evidence-based answers from a database of 200 million peer-reviewed scientific papers in Semantic Scholar and other sources. Consensus uses advanced AI, including GPT-4, to deliver concise summaries, extract key insights, and support natural language queries. Its hybrid search approach combines semantic searching (AI embeddings) and keyword search (BM25) to ensure relevance and precision, ranking results based on citation count, recency, and journal impact.
The Consensus Meter visualizes agreement or disagreement for yes/no questions; librarians can show it to researchers to get their input. There is also Consensus' Copilot, a virtual research assistant that drafts literature reviews, explains complex concepts, and provides citations; it answers questions, drafts content, creates lists, and more with mixed results. Users filter results by study design, publication type, or citation count, and save searches or papers for future reference. The platform integrates with tools such as ChatGPT enhancing academic writing with accurate citations.
New in Consensus 2025
In October 2025, Consensus announced its new "Pro Search and Scholar Agent" to improve search results. Similar to Undermind.ai, this new feature in Consensus aims to:
Run multiple complex searches in one command
Synthesize findings across topics into unified reports
Filter and reason through evidence like a real researcher
Leverage full-text papers for deeper, more transparent analysis
Consensus calls itself the best AI academic search engine for scientific and academic research. Its AI feature called “Copilot” enhances the user’s searching experience. Consensus Copilot will generate summaries with real citations (a nod to GenAI's tendency to hallucinate references).
Articles are labelled based on study type (e.g., systematic review, RCT) and journal reputation, allowing users to prioritize their reading;
“Study Snapshot” identifies a study's methods, extracting information such as sample size, population, and more.
Original text is extracted from papers, making it easier for users to verify information.
For Yes/No questions, the Consensus Meter displays a distribution of supporting and contrasting evidence.
Users have the option to limit the publication year range.
Users can add additional instruction to the prompt e.g. grouping evidence on both sides.
Librarian criticism
In its documentation, Consensus has a page addressing ethical AI use, which may be unique among AI companies.
Consensus aims to provide reliable, evidence-based responses to questions using citable research to minimize hallucinations. It succeeds in some cases where the evidence is strong but there are errors. Consensus says it is strong in medical and social policy research, but supports other disciplines beyond medicine. I might recommend using it in concert with other free sources, such as PubMed and Google Scholar, to test and verify. When I tested Consensus, I liked the presentation of information but the system revealed its limitations in handling complex queries and oversimplification in its summaries. Its user-friendly interface, advanced filters, and features like Study Snapshots and Ask Paper, are all welcome. Human researcher and "medical librarian in the loop" evaluations are needed. For more information, see Jarry 2025.
Consensus does not replace traditional bibliographic databases, though for some scholars not affiliated with a university, it may reduce time and resources allocated to searching. Academics and researchers are advised to verify any summaries generated by AI and ensure that the original articles are summarized accurately. For the public, Consensus offers a freemium model of 10 free AI-generated summaries per month.
Note: This presentation was selected by a librarian as an introductory video. However, as this is a marketing video and tutorial, some of the claims of the video should be tested and verified. See also https://www.youtube.com/watch?v=TKoqayujF_o
"... This research evaluates the performance of platforms such as SciSpace, Elicit, ResearchRabbit, Scite.ai, Consensus, Claude.ai, ChatGPT, Google Gemini, Perplexity, and Microsoft Co-Pilot across the key stages of SLRs—planning, conducting, and reporting. While these tools significantly enhance workflow efficiency and accuracy, challenges remain, including variability in result quality, limited access to advanced features in free-tier versions, and the necessity for human oversight to validate outputs..."
Note: Please use your critical reading skills while reading entries. No warranties, implied or actual, are granted for any health or medical search or AI information obtained while using these pages. Check with your librarian for more contextual, accurate information.
