Consensus
6.1/10Consensus is an AI search engine for peer-reviewed research. It finds, summarizes, and synthesizes scientific papers for students, researchers, clinicians, and research teams.
Data last updated November 2026 · Our methodology
Overview
Consensus is an AI academic search engine. It answers questions from a corpus of more than 200 million scientific papers instead of the open web. On yes or no questions, its Consensus Meter shows how many studies support, contradict, or are inconclusive on the claim. Study Snapshots summarize each paper's population, sample size, methods, and limitations. Deep Search spends several minutes reading many papers and writes a report that includes research gaps. A multi-agent Scholar Agent built on GPT-5 handles planning, searching, reading, and analysis. It suits students, academics, doctors, and evidence-focused teams in consultancies or healthcare who need cited sources more than general chat answers. Pricing starts with a free tier that has monthly quotas on premium analyses. Paid individual plans are Premium and Pro, then a per-seat Teams plan billed annually and a custom Enterprise tier for universities and large organizations. Citations export to Zotero, Mendeley, EndNote, and RefWorks. An MCP server lets clients such as Claude, Cursor, and ChatGPT query the corpus. A public API charges per call beyond the credits included in a plan. Enterprise plans support institutional login through SAML and Shibboleth. The main limitation is that the synthesis can oversimplify contested debates. The corpus also includes preprints, and there is no automatic filter for retracted papers, so users still have to read the key papers themselves. Trustpilot reviewers also complain about annual renewal charges after trials.
Our Rating
Who Consensus is best for
Good fit
It is built around searching and summarizing peer-reviewed literature for students and academics, and offers enterprise licensing for universities.
Clinicians can check the evidence behind a claim with the Consensus Meter and Study Snapshots.
Less suited
Useful for evidence-backed research, but it covers only academic literature.
The API and MCP server allow developers to build on the paper corpus.
Pros & Cons
Pros
- The Consensus Meter shows at a glance whether studies agree on a yes or no research question.
- Answers come only from academic literature, a corpus of more than 200 million papers, so every claim is cited to a paper.
- The free tier includes unlimited searches, and paid individual plans start at about 9 USD a month.
- Citations export to Zotero, Mendeley, EndNote, and RefWorks on every plan, including Free.
- An MCP server and a public API let teams use the corpus from Claude, Cursor, or their own tools.
Cons
- Its syntheses can oversimplify contested scientific debates.
- Preprints are included, and there is no automatic filter for retracted papers.
- The interface and synthesis are mainly in English.
- Trustpilot reviewers report unexpected annual renewal charges and trouble getting refunds.
- Covers academic papers only, not your internal documents or the open web.
Key Features
Consensus Pricing
Free
- Consensus Meter
- Unlimited searches
- 10 Pro Analyses per month
- 10 Study Snapshots per month
- Citation export to reference managers
Pro
or $120/year (save 33%)
- MCP access
- Scholar Agent Deep mode
- 500 API credits included
- 15 Deep Searches per month
Enterprise
- 500 API credits included
- SAML and Shibboleth institutional login
- Custom pricing for universities and large organizations
Teams
or $120/user/year
- Shared libraries
- 500 API credits included
- Centralized access management
- Billed annually at 9.99 USD per seat per month
Frequently Asked Questions
Yes. The free tier includes unlimited searches and the Consensus Meter, plus 10 Pro Analyses and 10 Study Snapshots per month. The quotas reset monthly.
Yes. Consensus says its API is open to everyone. Pro, Team, and Enterprise plans include 500 credits, and additional calls cost 0.05 USD each.
Yes. Exports to Zotero, Mendeley, EndNote, and RefWorks are available on all plans.
Enterprise plans support SAML and Shibboleth institutional login.
Yes. An MCP server lets MCP-compatible clients such as Claude Desktop, Claude Code, Cursor, and ChatGPT query the corpus. A third-party guide says this requires the Pro plan or higher.
Yes. Several Trustpilot reviewers say they were charged for an annual plan after a trial, so check your billing settings before the trial ends.
Similar Tools
See all alternatives
Amazon's public cloud platform, with over 200 pay-as-you-go services for compute, storage, databases, networking and AI, used by startups through large enterprises and governments.
Anthropic's organization-wide Claude plan adds chat, Claude Code, Cowork agents, connectors, and enterprise governance such as SCIM, audit logs, and a compliance API.
Fin is Intercom's AI customer service agent. It resolves support conversations across chat, email, voice and social, either inside Intercom or on top of an existing helpdesk.