Consensus
Visit Consensus →About Consensus
Consensus is a search tool built for academic and scientific literature. You ask a question in plain language and it pulls relevant peer-reviewed papers, summarizes what they found, and links directly back to the sources so you can check the claims yourself. It draws on a corpus cited at over 220 million papers.
The product has changed a lot since its early days. It launched around the idea of a "consensus meter" that showed the percentage of studies agreeing with a claim, and while that feature still exists, it is no longer the headline. The 2026 version (no numbered release) is organized around Deep Search and a Research Agent.
Deep Search breaks your question into sub-questions, runs up to roughly 20 targeted searches, and synthesizes across as many as 50 papers, with a wide-coverage mode that can scan up to about 1,000. The Research Agent handles multi-step research questions, Medical Mode restricts sources to clinical guidelines and top medical journals, and an improved MCP lets you use Consensus inside ChatGPT, Codex, Claude, and Claude Code.
Pricing is freemium. Free ($0) gives you basic Quick Search, abstract summaries, limited Pro messages, and about 3 Deep Searches a month. Student is $9/mo with .edu/.ac verification and up to ~40% off annual. Pro is $10/mo for unlimited Pro search, a larger Deep Search allotment, Study Snapshots, and full export (~33% off annual). A higher Deep tier is reported at $45/user/mo with around 200 Deep Searches a month and unlimited snapshots.
The limits are worth knowing. Free Deep Search is capped tightly, so heavy users will hit the paywall quickly. Coverage skews toward what is indexed and abstract-level for some records, and like any AI summarizer it can flatten nuance, so you still need to read the primary sources for anything high-stakes. Note also that an unrelated sales-demo product at goconsensus.com shares the name and costs $600-$1,250/mo; it is not this tool.
It suits researchers, students, clinicians, and analysts who want grounded, citable answers from the literature and are comfortable verifying against the originals rather than trusting a summary at face value.
- Answers are grounded in peer-reviewed papers with direct links back to sources, so claims are checkable
- Deep Search decomposes a question and synthesizes across dozens of papers (up to ~1,000 in wide mode)
- Very large corpus cited at 220M+ papers
- Medical Mode narrows to clinical guidelines and top journals for health questions
- MCP integration lets you use it inside ChatGPT, Claude, Codex, and Claude Code
- Free plan caps Deep Search at about 3 per month, so serious use needs a paid tier
- The higher Deep tier at $45/user/mo is a jump for individuals
- AI summaries can lose nuance; primary sources still need reading for high-stakes work
- Coverage is limited to indexed literature and some records are abstract-level only
- Name collision with an unrelated sales tool at goconsensus.com can cause confusion
Choose Consensus if…
- You need evidence-backed answers with citations from academic literature
- You run multi-step or literature-review-style research and want it partly automated
- You work in medicine and want answers limited to guidelines and top journals
- You want to query the scientific literature from inside ChatGPT or Claude
- You are a student or academic who can use the $9-$10/mo tiers
Leave a Review
Reviews are published after moderation. We don't share your email.