Consensus
AI search engine across 400M research papers
- Overview
- Consensus searches over 400 million peer-reviewed papers and returns cited, evidence-grounded answers in plain English. Its Consensus Meter visualizes scientific agreement across studies for yes/no questions, while Deep Search runs multi-paper literature reviews across up to 100 papers in a single query. The platform integrates with ChatGPT, Claude, Microsoft 365 Copilot, Zotero, and a public API.
- Why it matters
- Consensus is one of the few AI research tools grounded entirely in peer-reviewed literature rather than open-web retrieval, which matters as hallucination risk grows in general-purpose AI assistants. Adoption is expanding from individual academics into clinical and policy teams that need defensible, traceable evidence. The mid-2026 API launch and Microsoft 365 Copilot integration signal a push into enterprise workflows, where the key risk to watch is per-seat pricing creep on Teams contracts and reliance on a proprietary paper corpus that competitors cannot replicate.
- In practice
- Research and medical affairs teams use Consensus to run structured literature scans before drafting white papers or clinical protocols — the Deep Search tier handles the initial 50–100 paper sweep that would otherwise take days. The MCP server integration lets developers pipe Consensus results directly into internal AI agents or coding tools without manual copy-paste. The honest limitation is that Deep Search is capped monthly by plan, so heavy users running multiple systematic reviews per week will hit the 200-review ceiling on the Deep plan and face custom enterprise pricing.
- Best for
- Researchers, clinicians, and analysts who need citation-backed evidence synthesis without manually reading dozens of papers
- Pricing
- Free (10 Pro messages/mo, 3 Deep reviews/mo); Pro $20/mo or $144/yr (unlimited Pro messages, 15 Deep reviews/mo); Deep $65/mo or $540/yr (200 Deep reviews/mo); Teams and Enterprise custom
Key features
- Consensus Meter showing scientific agreement level across studies
- Deep Search: multi-agent literature review across 50–100 papers per query
- Research Agent for multi-step, agentic research question planning
- Self-serve API and MCP server for embedding search in external tools
- In-results figures and tables extracted from cited papers
Quick answers
- What is Consensus?
- Consensus searches over 400 million peer-reviewed papers and returns cited, evidence-grounded answers in plain English. Its Consensus Meter visualizes scientific agreement across studies for yes/no questions, while Deep Search runs multi-paper literature reviews across up to 100 papers in a single query. The platform integrates with ChatGPT, Claude, Microsoft 365 Copilot, Zotero, and a public API.
- Why does Consensus matter?
- Consensus is one of the few AI research tools grounded entirely in peer-reviewed literature rather than open-web retrieval, which matters as hallucination risk grows in general-purpose AI assistants. Adoption is expanding from individual academics into clinical and policy teams that need defensible, traceable evidence. The mid-2026 API launch and Microsoft 365 Copilot integration signal a push into enterprise workflows, where the key risk to watch is per-seat pricing creep on Teams contracts and reliance on a proprietary paper corpus that competitors cannot replicate.
- What is Consensus best for?
- Researchers, clinicians, and analysts who need citation-backed evidence synthesis without manually reading dozens of papers
- How much does Consensus cost?
- Free (10 Pro messages/mo, 3 Deep reviews/mo); Pro $20/mo or $144/yr (unlimited Pro messages, 15 Deep reviews/mo); Deep $65/mo or $540/yr (200 Deep reviews/mo); Teams and Enterprise custom
Seen in recent stories
Where Consensus showed up in the last 90 days of KeyNews editions.
Know the terms
Glossary entries that explain what Consensus is built on.
RAG (Retrieval-Augmented Generation)
A technique that retrieves relevant documents from an external knowledge base and feeds them to a model alongside the user's query. RAG reduces hallucination and keeps responses grounded in current, factual data.
Hallucination
When an AI model generates confident-sounding but factually incorrect or fabricated information. Hallucination is the number-one barrier to enterprise AI adoption and a major focus of current research.
Agentic RAG
A retrieval-augmented generation system where an AI agent autonomously decides when, what, and how to retrieve information — dynamically choosing between multiple knowledge sources, reformulating queries, and iterating on results rather than making a single fixed retrieval call.
Grounding
Techniques that anchor AI outputs in verifiable facts by connecting models to external knowledge sources. Grounding reduces hallucination and is essential for enterprise use cases where accuracy is non-negotiable.
New tools drop every week.
ONE BRIEFING · EVERY FRIDAY · FREE
Free. Unsubscribe anytime.