Write a Research Paper in No Time with Consensus AI | Writing literature review using Consensus

Consensus AI is positioned as an academic research assistant that turns literature-search results into structured, citation-backed answers—aiming to speed up tasks like literature reviews and evidence synthesis without requiring users to craft complex prompts. After logging in (including via Google or Facebook), users can search across a wide range of topics—mental health, business, healthcare, computer science, education, psychology, economics, and more—while the system emphasizes that its outputs are grounded in “top research articles” rather than generic web results.

A core feature is the “consensus meter,” which quantifies how studies lean toward different outcomes. In a demonstration question—whether online mental health counseling services are effective—Consensus aggregated findings from 13 relevant papers and produced a distribution: 62% of studies favor “yes,” 0% favor “no,” 8% suggest “possibly,” and 31% show mixed results. The answer then comes with a detailed breakdown organized under headings such as effectiveness, factors influencing effectiveness, and comparisons with traditional counseling. Each claim is tied to specific citations; hovering over numbers reveals the underlying sources, and the full reference list appears at the end.

For literature reviews, the workflow is framed as simple and prompt-light: users write a topic in their own words (example: “impact of CSR on effective commitment”), and Consensus generates an introduction, thematic sections, mechanisms, cross-cultural considerations, and a conclusion based on a “pro analysis of top 10 papers.” The output is presented as text that can be copied either without citations or with in-text citations plus a complete references list. Users can also choose different citation styles (including APA, Harvard, and Chicago) and export references in RIS format for use in tools like EndNote or other citation managers.

The transcript also stresses guardrails. Copying the generated literature review verbatim is discouraged; users still must write in their own words. A cautionary anecdote claims that an external examiner found more than 95% of citations in an undergraduate thesis to be fake, suggesting misuse of AI tools like ChatGPT for literature reviews. Consensus is presented as a mitigation because it is connected to a large corpus (described as 200 million research articles), enabling citation-linked outputs rather than fabricated references.

To refine evidence, Consensus includes filters such as publication year (e.g., after 2020), open access, “ask paper available” (chatting with open-access papers), minimum citation counts, research method categories (meta-analysis, systematic review, sample-size-related options), journal tiers (Q1/Q2), domain, and country. After applying filters, the system rewrites the literature review accordingly and provides the final set of references used.

Additional features include “study snapshots,” which summarize each study’s population, setting, method, outcomes, and key results—intended to replace manual fact-sheet extraction. Pricing is described as offering a free tier with unlimited searches and quality indicators, and a premium tier adding GPT-4-powered pro analysis, study snapshots, paper messaging, bookmarks, custom lists, and automatic updates. Students are told they can claim a 40% discount using a verified .edu or .ac email address.