Consensus

Consensus AI is an AI-powered academic search and research platform designed to help students, researchers, clinicians, academics, and professionals find, understand, and synthesize scientific literature more efficiently. Rather than functioning as a general-purpose chatbot that answers questions from broad model knowledge, Consensus searches a large academic research database first and grounds its AI-generated responses in research papers. Its current database contains more than 220 million peer-reviewed research papers, with coverage across scientific and academic disciplines.

The platform is built around a research-first workflow. A user can enter a natural-language question, a topic, keywords, a Boolean query, a paper title, an author, or a DOI. Consensus then searches its research corpus and ranks relevant papers using a combination of semantic search and keyword matching. It also considers signals such as publication recency, citation count, and journal quality when refining results. This approach allows researchers to search for evidence using ordinary questions rather than relying entirely on carefully constructed academic database queries.

One of Consensus’s most recognizable features is the Consensus Meter. When users ask a suitable yes-or-no research question, the system analyzes the conclusions of the most relevant papers and presents a visual indication of whether the literature leans toward Yes, No, Possibly, or Mixed. The tool is intended to provide a quick overview of agreement or disagreement within the research rather than treating a single paper as definitive. Consensus requires at least five relevant papers before displaying the Meter.

The platform also includes Pro Search, which provides an AI-generated synthesis of the research returned for a question. Consensus says Pro can analyze up to 20 papers and provide summaries, key findings, and connections between studies. Every AI-generated synthesis includes citations so users can trace claims back to the underlying research. This makes Pro particularly useful for quickly understanding a research area before reading individual papers in greater depth.

For more demanding research projects, Consensus offers Deep Review. Deep can conduct a substantially broader literature review across approximately 50 to 100 papers, depending on the workflow. Instead of providing only a quick summary, it is designed for more comprehensive exploration of complex questions and can compare evidence, identify themes, and synthesize findings across a larger collection of studies. This makes Deep more relevant to literature reviews and in-depth research than a simple paper search.

Consensus has also introduced a Research Agent that can handle complex, multi-step research questions. The agent can plan research steps, chain together different tools, apply academic filters, search for related papers, compare studies, perform DOI and author lookups, explore citation relationships, and identify potential gaps in the literature. Its responses are grounded in peer-reviewed research and include citations. This turns Consensus from a conventional academic search engine into a more agentic research environment.

Another useful feature is Study Snapshot. Users can open a paper and quickly extract important study details such as methodology, sample size, study duration, outcomes, and other research characteristics. This can save researchers from repeatedly reading full papers simply to determine whether a study uses a relevant method or population. Study Snapshots are available without limits on the Pro plan and are also included in the Free tier in a limited quantity.

Consensus also provides Chat with Full Text, which allows users to ask questions about the complete text of specific papers, collections, or uploaded documents. This is particularly valuable when a researcher needs to investigate methods, results, limitations, or other details that may not be fully represented in an abstract. Consensus states that where full text is available, its search system can go beyond titles and abstracts and work with sections such as methods, results, and discussion.

A related feature is My Library, which allows users to save papers and organize them into collections. Researchers can upload their own PDFs, import references, and work with saved papers in a more structured environment. Collections can be searched and analyzed independently, helping users maintain project-specific research libraries instead of repeatedly running searches across the entire Consensus database.

Consensus also provides a Citation Graph for visualizing relationships between research papers. The current system supports graphs containing 20, 40, or 60 papers and can be used to explore clusters and citation relationships around a research topic. This can help researchers identify influential papers, connected studies, and additional literature that may not appear in a simple keyword search.

The platform supports several research-management workflows. Search results can be exported as CSV or RIS files, allowing users to move references into tools such as Zotero, EndNote, Mendeley, Paperpile, RefWorks, Citavi, and Papers. Consensus also supports importing from Zotero collections and working with DOI, RIS, and BibTeX information. This allows the platform to fit into established academic reference-management workflows without attempting to replace every reference manager.

Consensus has expanded its availability through ChatGPT and Claude. Its MCP server enables AI assistants to search the Consensus research database directly, giving users access to the platform’s academic corpus from within supported AI environments. The current MCP documentation specifically supports ChatGPT, Claude, Copilot, and other MCP-compatible clients. Users can therefore combine Consensus’s research index with conversational AI workflows while maintaining research-focused retrieval.

The platform also provides a public API. The API allows developers to programmatically search academic research, retrieve relevant papers, and integrate Consensus search into their own applications and workflows. API usage shares its monthly allowance with MCP usage, and calls can be filtered by publication date, study type, sample size, citation count, journal, academic field, country, open-access status, and other parameters. This makes Consensus suitable for custom research applications and automated evidence-gathering workflows.

Consensus is also designed to work with institutional libraries. Users can select their university or institution in account settings so that Consensus can provide more accurate full-text links through library access systems such as LibKey. The platform also offers enterprise and institutional plans with library integrations, centralized management, and support for larger research communities.

The platform covers a broad range of scientific disciplines. Its current documentation says the database includes research across fields ranging from astronomy and sociology to biochemistry and includes all of PubMed. A dedicated Medical mode searches a curated subset of approximately eight million papers and 50,000 clinical guidelines from leading medical journals. This gives healthcare and clinical researchers a more focused environment when exploring medical evidence.

Consensus is strongly oriented toward evidence-based research, but it does not present its AI outputs as a replacement for reading the original literature. The platform emphasizes citations and source traceability, and its own research methodology says it will indicate when it cannot find sufficient relevant evidence instead of filling gaps with unsupported outside information. This is particularly important when using AI for scientific or academic work, where conclusions need to be checked against the actual studies.

The company was founded in 2021 by Christian Salem and Eric Olson, who created Consensus with the goal of making scientific knowledge easier to find, understand, and trust. The company began as a remote-first organization and is now based in San Francisco. Its current website reports that more than 10 million researchers, students, and clinicians use Consensus and that its platform has handled more than 150 million research questions. These are company-reported usage figures.

Consensus follows a freemium subscription model. The Free tier includes unlimited paper searches, limited Pro messages, three Deep reviews per month, 10 Study Snapshots per month, and 30 API/MCP calls. Pro costs $20/month or $144/year, equivalent to $12/month when billed annually, and provides unlimited Pro messages, 15 Deep reviews, unlimited Study Snapshots, and 250 API/MCP calls per month. A Deep plan is available at $65/month or $540/year, with 200 Deep reviews and 1,000 API/MCP calls per month. Teams use custom pricing, while Enterprise pricing is also custom.

Consensus does not currently have a conventional free trial separate from its Free tier. Users can begin with the free version and upgrade when they need higher limits. Paid API/MCP usage beyond the included allowance can be enabled at $0.05 per additional call, subject to the limits associated with each plan.

Independent review information requires careful handling because another company called Consensus operates at goconsensus.com and has thousands of reviews for sales-demo software. Those reviews do not belong to the academic research product at consensus.app and should not be used for this listing. For the academic Consensus product itself, public review volume is comparatively limited, while recent Trustpilot feedback for consensus.app is mixed, with a 2.5/5 TrustScore from 6 reviews as of the latest check. Several recent reviews criticize billing experiences, support responsiveness, or research retrieval limitations. Because the Trustpilot sample is very small, it should be treated as a warning signal rather than a comprehensive measure of product quality.

Consensus is best suited to students, researchers, academics, clinicians, universities, and evidence-focused professionals who need to find and synthesize peer-reviewed research quickly. Its strongest differentiators are its large academic corpus, citation-backed AI synthesis, Research Agent, Consensus Meter, Deep Reviews, Study Snapshots, full-text analysis, My Library, Citation Graph, API, and MCP connectivity. It is less appropriate as a general writing assistant, marketing platform, CRM, or content-generation suite.

Pricing

• Free: $0. Includes unlimited paper searches, limited Pro messages, 3 Deep reviews/month, 10 Study Snapshots/month, and 30 API/MCP calls/month.
• Pro: $20/month or $144/year, equivalent to $12/month when billed annually. Includes unlimited Pro messages, 15 Deep reviews/month, unlimited Study Snapshots, and 250 API/MCP calls/month.
• Deep: $65/month or $540/year, equivalent to $45/month when billed annually. Includes 200 Deep reviews/month and 1,000 API/MCP calls/month.
• Teams: Custom pricing.
• Enterprise: Custom pricing.

Review

The academic research product at consensus.app has limited independent review coverage. Recent Trustpilot data shows a 2.5/5 TrustScore from 6 reviews, with several recent complaints involving subscription/billing experiences and support. Because the sample is very small, this rating should not be treated as representative of the entire user base.

It is important not to confuse consensus.app with the separate Consensus sales-demo platform at goconsensus.com, which has a much larger G2 and Capterra review base. Those ratings belong to a different product.