カオスマップに戻る

Overview

Consensus is an AI search engine focused on "visualizing scientific consensus." It covers 220+ million peer-reviewed academic papers and features the Consensus Meter, which shows how many papers answer a question with Yes, No, or Possibly. Unlike generative chat AI, Consensus is built for users who need evidence-based answers grounded in real research.

Key Features

1. Consensus Meter

Analyzes relevant papers and visually displays the percentage supporting Yes / No / Possibly for a given claim. Especially valuable for contested fields like medicine, nutrition, and psychology.

2. 220M+ Peer-Reviewed Papers

The corpus is built on OpenAlex and Semantic Scholar and supplemented by full-text partnerships with Taylor & Francis, Sage, and the American Chemical Society. All answers come from peer-reviewed literature with direct citation links — unlike open-web search AI, Consensus prioritizes primary source credibility.

3. Research Agent

The successor to Scholar Agent. It plans its own searches for complex multi-step questions, chains tools together, applies academic filters, and returns citation-backed answers grounded in peer-reviewed research.

4. Deep Search

Automates literature reviews by building search strategies, expanding key terms, identifying conflicting arguments, and exploring citation graphs, producing a structured report in minutes.

5. Result Filters

Narrow results with 19 study-design filters grouped into Experimental, Observational, Evidence Synthesis, and Other categories, plus 44 publisher filters.

6. Research Hub / Ask Paper

Upload a PDF and chat with it directly, pinpointing answers inside long papers. Clicking a search result opens a paper preview in a sidebar.

7. Medical Mode

A specialized mode that narrows results to clinical guidelines and top medical journals — strong support for healthcare professionals making clinical decisions.

8. External Clients

Consensus is available as a ChatGPT App and connects to Claude over MCP.

Pricing

PlanMonthlyKey Perks
Free$0Unlimited basic search, Pro features capped
Premium$8.99 (annual)Pro Analysis, Study Snapshots
Pro$15 ($120/yr)15 Deep Searches per month
Deep$45Unlimited Deep Search for high-frequency reviews
Teams$9.99/seatTeam usage
EnterpriseContact salesOrg-level features and admin

The API is billed at $0.10 per call plus a platform fee.

Typical Use Cases

  • Validating medical, nutrition, or psychology claims against real papers
  • Fact-checking for writers and journalists
  • Evidence gathering for policy and risk assessment
  • Clinical decision support for healthcare professionals

Strengths

  • Extremely credible: 220M+ peer-reviewed papers only
  • Consensus Meter summarizes scientific agreement at a glance
  • Medical Mode tailored for clinicians
  • Research Agent automates multi-step investigations

Weaknesses

  • Does not cover general news or blog content
  • Very recent research may take time to appear
  • Non-English (e.g. Japanese) literature coverage is weaker

Official

公式サイト