> ## Documentation Index
> Fetch the complete documentation index at: https://docs.consensus.app/llms.txt
> Use this file to discover all available pages before exploring further.

# What is Consensus

> Consensus is an AI search engine for research. It searches 220 million peer-reviewed papers, analyzes the results, and gives you cited answers you can trace back to the source.

<iframe width="100%" height="400" src="https://www.youtube.com/embed/qhiIG0y-tko" title="Consensus is now a Research Agent (not just a search tool)" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen />

Consensus is an AI-powered academic search engine built on a database of over 220 million peer-reviewed research papers. Unlike general AI tools, every response is tied back to a real research paper. Think of it as an AI-native alternative to Google Scholar with a more refined corpus.

Consensus is a search engine, not a chatbot. For every query it first retrieves the most relevant papers, then uses AI to analyze the top results and synthesize the findings. Every response includes citations, so each insight is one click from the original source.

Over 10 million researchers, clinicians, students, and professionals use it. It's free to start.

## What you can do with it

* **Answer a research question** with a cited synthesis across the strongest papers, plus a Consensus Meter showing where the evidence leans on yes/no questions.
* **Go deeper on hard questions** with [Deep Search](/core-features/deep-search), a longer, more robust version of a search that runs up to 20 searches in parallel and reads 50 to 100 papers.
* **Chain research tasks together** with [Research Agent](/core-features/research-agent): citation crawling, author search, comparison tables, gap analysis, and filters stated in plain language.
* **Read the full paper, not just the abstract.** Through publisher partnerships and open-access PDFs, Consensus analyzes methods, results, and discussion sections. See [Full text access](/get-started/full-text-access).
* **Organize and search your own papers** in [Library](/core-features/library), including uploads and Zotero imports.
* **Explore how papers connect** with the [Citation Graph](/core-features/citation-graph).

## Three ways to use Consensus

<CardGroup cols={3}>
  <Card title="Web app" icon="globe" href="https://consensus.app">
    Search, Deep Search, Library, and Citation Graph at consensus.app.
  </Card>

  <Card title="MCP" icon="plug" href="/consensus-mcp">
    Search from inside Claude, ChatGPT, Cursor, and any MCP client.
  </Card>

  <Card title="API" icon="code" href="/api-get-started">
    One REST endpoint for your own apps, agents, and pipelines.
  </Card>
</CardGroup>

## Search modes

| Mode         | What it does                                                                                                           | Papers analyzed |
| ------------ | ---------------------------------------------------------------------------------------------------------------------- | --------------- |
| Paper search | A ranked list of relevant papers, no AI summary                                                                        | —               |
| Pro          | A cited summary with connections between findings. The default.                                                        | Up to 20        |
| Deep         | A longer, more robust version of Pro: up to 20 parallel searches and multiple Research Agent tools like citation crawl | 50–100          |

Switch the corpus between **All** (220M+ papers), **Medical mode** (\~8M papers and 50,000 clinical guidelines from the top 1,000 medical journals), and [**My Library**](/core-features/library).

## How Consensus removes hallucinations

* **AI runs after search, not instead of it.** Models only summarize the papers that were retrieved; they don't fill in blanks with outside knowledge.
* **Relevance checkers gate every answer.** Separate models verify that a paper actually contains information relevant to the query before it's used. If nothing relevant is found, Consensus says so.
* **Every statement is traceable.** Hover a citation to see exactly where in the paper it comes from, then click through to the passage. See [Citation Grounding](/core-features/citation-grounding).
* **Retracted papers are flagged** with a RETRACTED badge and never used in analyses.
* **Your data isn't training data.** Consensus doesn't use your data to train models or share it with third parties. Uploads are private to your account.

## Plans

| Plan                   | Pro messages | Deep searches         | API & MCP calls        |
| ---------------------- | ------------ | --------------------- | ---------------------- |
| Free                   | 10 per month | 3 per month           | 30 per month           |
| Pro ($20/mo, $144/yr)  | Unlimited    | 15 per month          | 500 per month          |
| Deep ($65/mo, $540/yr) | Unlimited    | 200 per month         | 2,000 per month        |
| Teams                  | Unlimited    | 50 per user per month | 500 per user per month |

Paper search is unlimited on every plan. See [pricing](https://consensus.app/pricing) for details, and [help.consensus.app](https://help.consensus.app) for account help.

## Next

<CardGroup cols={2}>
  <Card title="Where do papers come from" icon="database" href="/get-started/where-papers-come-from">
    The sources behind the 220M-paper index.
  </Card>

  <Card title="How search works" icon="search" href="/get-started/how-search-works">
    Hybrid retrieval, quality re-ranking, and grounded synthesis.
  </Card>
</CardGroup>