The way we extract information for literature reviews has completely changed in the age of AI. Consensus is an AI-driven search and literature review tool designed to speed up the research process and find more important information faster. It scans millions of papers to find the ones that match your query and sort them by which support or reject your argument. In this review, we’ll break down the main use cases of Consensus, its limitations and key features to make your research process more productive. We’ll also look at ConsensusGPT, which combines ChatGPT’s conversational capabilities with Consensus’ expansive database.
Learn which tools can speed up your research: This free 21-day email series introduces you to a range of academic tools and how they work together to speed up the time needed for a literature review. Most of these tools have been around for less than a few years, some use AI. Knowing them gives you an almost unfair productivity advantage for your research.
What problem does Consensus solve?
Literature reviews are notoriously challenging. They require in-depth review of dozens, or sometimes hundreds of research articles. It’s difficult to identify the most relevant papers, quickly synthesise them, and to infer the state of the field based on the literature. Consensus AI aims to accelerate all of these steps by immediately locating the most relevant papers and synthesising their insights in a useful way, so you can quickly understand your field and make progress in the lit review.
Ultimately, this means you can can learn in seconds, rather than hours, what works and what doesn’t for a potential research questions or direction.
Below, we’ll dive into the best features Consensus offers and how best to use the tool.
How to Use Consensus AI: Step-by Step Guide
2026 UpdateConsensus has expanded considerably in 2026. Namely, its new Research Agent can plan and execute multi-step research tasks; there is a new and improved research workspace; it now supports Zotero imports, collections, bibliography generation and more; Deep Search can also analyse saved or uploaded collections of up to 50 papers; Consensus now has broader access to full-text research (ie. not just abstracts) from major academic publishers.
Getting started with Consensus is as simple as typing a question into a textbox. It’s free when getting started. However, it is crucial to understand this technology’s limitations and know which results you can fully trust and which you need to double-check. We’ll cover these details as we review Consensus in detail below.

To get started with Consensus, go to the site and sign up for a free account. Then, start a new search using a research query. You can use your research question, a more specific query or simply keywords on a topic. By default, it’ll run your search as a “Pro Search”. You can also run Deep Searches, which we review below in detail. What works especially well is asking yes/no questions.
Consensus will return results like this:
- Consensus of the literature: What % of papers answer your question as yes, no, possibly or mixed.
- Summary with citations
- Key findings with corresponding citations
- Full list of references

At the end of your search you will find a number of studies sorted by relevance. Click each of these to see the detailed study view. This view brings together all the different pieces of information on this study which includes metadata, summaries, quality indicators, citation counts and a link to the publication itself. You can view any study’s abstract or snapshot; and if the paper is Open Access, you can see the PDF in the app too. However, PDF viewing within Consensus can slow down the app and put a lot of strain on your CPU. This was the case for me, using Consensus on Google Chrome.

How to use Consensus: Prompt Tips
To get the most out of Consensus, ask these kind of questions:
✅ Yes / No
✅ Relationship (e.g. “Is Sauna beneficial for blood pressure”)
✅ Benefits of… (e.g. “Benefit of sauna”)
Open-ended questions or ones that require numbers/explanations may or may not work. This is a general case with AI. Questions requiring logical reasoning and thinking rarely produce excellent results (as of 2025).

For example, this is not a good question: “How much climate change can European forests tolerate?”. On the other hand, converting it to a relationship-based question can provide reliable results, like “What factors determine European forests resilience to climate change?”.
While you will almost always get a reply from the AI, the value will diminish the more you deviate from questions the AI can effectively answer. This is not a particular limitation of Consensus AI, but applies to all large language models (think GPT, Claude, Gemini, etc.) since they deal with facts as language rather than logically connected entities.
You can check out the much more detailed guide on the Consensus FAQ page. The best way to find out, however is to just try it out for yourself.
The Consensus Meter for yes/no Questions
When asking Yes/No questions, Consensus attempts to give you a general consensus of the literature and report it using The Consensus Meter. This consists of four bars: Yes, No, Possibly and Mixed.
Consensus AI first finds several papers related to your query, scans them, and then sorts them into one of the three categories. If most papers affirm your question, the yes bar will grow. This is why Consensus invites you to use yes/no questions. Otherwise, the meter would make little sense. You get 25 of these “Pro Searches” for free monthly.

This feature is handy even when researching what supplements to take or what exercise to prioritise. In my experience, I tend to trust this reply over clickbait Google articles written primarily for SEO (Search Engine Optimisation) and “approved” by somebody with a doctoral degree, as their purpose is mainly to rank high in the Google search and only secondary to providing scientifically correct output.
How to do an Advanced Search with Consensus (Deep Search)
Consensus offers “Deep Search” as a way to get a more in-depth, automated literature review overview of a research question. It essentially conducts a fully automated and iterative literature review, step-by-step, showing each search along the way and the corresponding results.
Deep Search is far more computationally intense and takes significantly longer in time than a default search with Consensus. However, it’s far more robust and the results much more detailed. In fact, it’s akin to essentially running through an entire literature review with AI.

Results from Deep Search with Consensus
Based on a single research question, Consensus reviews the literature and builds out a detailed literature review report on your topic. The results include:
- Introduction: A detailed introduction, and overall yes/no breakdown (“Consensus Meter”)
- Methods: Context for the search strategy: total studies reviewed, screened, included, etc.
- Results: Answer broken down in different sections (evidence, patterns, mechanisms, etc.) as well as a Results Timeline with publications shown over time, and the top authors in the field.
- Discussion: Thoughtful discussion of answer, with useful Claims and Evidence table. This shows claims with corresponding evidence strength, reasoning and related papers.
- Conclusion: A final conclusion, with research gaps and a research gap matrix showing which areas are least explored, as well as related open research questions and why they matter.
Overall, the detailed output is exceptional, providing a wealth of knowledge on the subject, as well as corresponding references for every claim stated.
However, one drawback to the Consensus results is that the application doesn’t provide “deep links” into PDFs. For example, we can see a claim and the corresponding paper that Consensus uses to determine this claim. But we’d need to read the original source paper in detail to verify it’s findings. Other tools today now provide “deep links”, where the AI highlights the specific sentences within the PDF from which it took the information, making it easier to verify claims. Here’s how to do so in Sourcely, another lit review tool.
How to get study “snapshot” in Consensus
Consensus is handy both for the the big-picture and paper-specific information. It takes key pieces of information and metadata from articles to generate a clear “snapshot” of the paper. To view a study snapshot, click on “Study Snapshot” next to the “Abstract” option for any paper.

In the example above, you can see key pieces of information about the study, like:
- Population
- Location
- Methods
- Outcomes
- Results
It even reports how many of the total number of attributes it was able to pull (5/7 in this case).
This can be particularly useful in medical domains, because you can immediately see who the test subjects were, how many, which methods were used, and many other details that would likely take a while to find in the paper’s methods section.
Badges follow a similar idea but highlight different important properties of studies. For example, some studies might have more citations, while others employ a particular method. Badges make it easier to find papers that stand out.
Consensus also provides all the other essential information about articles on their detailed page’s, including title, author, abstract, citations, journal, quartile ranking and, of course, a link to the source.
How to filter in Consensus
Since Consensus is primarily a search engine, you can filter your results like on PubMed. In fact, Consensus is unique compared to other AI search tools in just how comprehensive its filtering options are. You can filter on:
- Publication Year
- Journal Rank
- Citation Count
- Methodology (including study types, human studies, populations, etc.)
- Field of Study
- Country
These filters get applied to the results used by Consensus to generate its results.

Many of these filters are useful across domains, and particularly for medical questions where sample size or human studies are essential distinguishing factors. I hope that one day Consensus may implement similar filters to search in my domain (ecology), for example, by study location, taxa, or statistical methods.

Consensus in ChatGPT
As of late 2026, Consensus is now available as a ChatGPT App, replacing the older Consensus GPT/plugin. Once connected, it lets ChatGPT search Consensus’s database of more than 220 million research papers directly inside a conversation. You can use it to find peer-reviewed studies, generate cited summaries, compare findings across papers, identify research gaps, and find supporting citations for your own writing.
To use Consensus in ChatGPT it, open the ChatGPT Plugins directory and search for Consensus and connect your Consensus account. It works on all ChatGPT plans, including the free one.

To then use Consensus in a conversation with ChatGPT, you need to mention it or “invoke it” by literally writing @Consensus anywhere in your message.
How to use Consensus in ChatGPT for Academic Writing
One excellent (and perhaps surprising) use case for Consensus is academic writing. I often use this workflow when writing a paper. Usually, I know certain things are true because I have read them many times but don’t have a specific citation in mind to back them up in a published paper. This is where you can use ConsensusGPT. Use the straightforward prompt “Find evidence for …”. You can use the results of several studies as a reference in your writing.
It goes without saying that you should always double check the reply by at least skimming the paper. Keep in mind that often AIs will not have access to the entire PDF but only the publicly available abstract and its conclusions might therefore be incomplete.

Consensus AI Reliability, Biases and Limitations
In this section, we’ll look at how accurate the Consensus responses are, and how best to use the tool given its limitations.
Is Consensus AI reliable?
There is an ongoing debate about how much we should trust AI. In my opinion, it’s not too different from asking a human. If you ask a five-year-old kid a serious question, you won’t seriously bet on the correctness. If they’re a 16 year old and their answer is surprising but not unrealistic, you may trust them but still double-check the answer. If a professor with 40 years of experience answers a question in their domain, their answer is as firm as it gets. The problem with AI is we don’t know which persona we are talking to.
My rule of thumb is to consider how much data has been published on a topic. If I am asking a question that has been debated for decades, there is likely a large amount of data available and I tend to trust AI (like in the example question above). The more recent and specialised the topic is, the less you can trust any AI.
Overall, the speed of AI is increasing very fast, and the training datasets are growing daily. Don’t hesitate to be specific and treat Consensus as a new colleague you know is knowledgeable but has yet to talk to them much. Get to know one another.
Will Consensus work in your subject area?
Research in certain scientific domains may be better suited for Consensus. According to the Consensus website, the best questions are in the medical and social policy domains. However, it is likely able to address many scientific disciplines because the database it relies on spans a large number of subjects. Consensus, like a lot of other AI research tools, relies on Semantic Scholar for data. Semantic Scholar has good coverage across disciplines; in fact, according to this study, it is second best only to Springer Link.
I have been using Consensus for my own research in forest ecology and it has always come up with reliable studies and answers. A good way to test this is to evaluate Consensus on a topic you’re already familiar with and ensure the search finds the essential papers you know are central to it.

Although Consensus AI is suited for a variety of subjects, the interface favors medical research. For example, you can filter by human/animal studies or find randomized control trials. These concepts don’t exist in most non-experimental or non-medical domains. However, the UI will likely change in the future as Consensus gets more adopted and expanded across all domains.
Limitations of Consensus AI
The two most significant limitations of all AI tools are incompleteness and irreproducibility. Incompleteness means that the results you get are filtered by an AI model with a certain level of randomness. We don’t truly understand how this process works, so it might miss some research that a human would deem necessary. This is especially true since we need to know which papers Consensus can access.
This brings us to the second issue: Irreproducibility, which is of particular concern when conducting a systematic literature review. Such a review involves documenting your paper search process, theoretically allowing for replication. However, if you were to state, for instance, ‘I searched for X on Consensus’, and someone attempted to replicate this a year later, their findings will likely differ from yours. This variation directly results from the stochastic nature inherent in all AI models.
Consensus AI vs ChatGPT
Consensus is fundamentally different from ChatGPT because of the data it uses to generate results. Consensus is expressly trained on academic texts, sourced mainly from Semantic Scholar, a large academic research database with over 200 million articles. ChatGPT, on the other hand, is trained largely on websites and books, and thus its responses are not as academically-focused. This is what makes Consensus, and other tools like it, so much more useful and applicable for researchers.
Consensus is one of many different AI tools for literature review. Others include SciSpace, Liner, Sourcely and Elicit. If you’re not sure where to begin, or which tool is best, I strongly suggest you take my free 21-day note-taking course. This will walk you through the most essential tools and workflows to maximise your productivity without compromising your academic integrity.
Consensus AI vs SciSpace
Both Scispace and Consensus have similar capabilities and the same goal of automating literature review using AI. While SciSpace shines at using AI to help you read papers and discover related papers, Consensus extracts essential information from a collection of documents. It displays the final results as a quick, easily digestible summary, or as a highly detailed report when using the Deep Search mode.
Although tools like SciSpace and NotebookLM (now Gemini Notebook) are equally helpful for analysing collections of literature, Consensus is entirely unique in how it answers Yes/No questions. As its name implies, it truly seeks to give you a pulse on the consensus of the literature, or the overall breakdown of how papers support or contend a research theory.
Compared to SciSpace, Consensus provides more customisations. In particular, it has a lengthy set of filters you can use to limit results, thereby focusing the summary and overview to papers of specific quality, population, study types, etc.
Overall though, both tools are useful for automating the discovery process in the initial stages of a literature review, as well as for discovering relevant literature and understanding your field.
👉 Learn more about SciSpace in my deep dive on SciSpace.
Consensus AI Pricing and Discount
Consensus AI is free, but also offers a paid tier with more advanced features. On the free plan, you can do up to 25 Pro Searches a month, which generate results using 20 papers; and up to 3 deep searches a month, which review up to 50 papers in detail. However, if you’d like to run more than this, you’ll need to upgrade to the premium tiers.
Use my discount code EFFORTLESS30 for a 30% discount. You may also qualify for other discounts too: students can purchase the premium membership at a 40% discount.

Is Consensus AI free?
Yes, Consensus AI is free. On the free tier, you have access to all the features but can only run a limited number of searches monthly. If you need to run more searches, you’ll need to upgrade to the Pro or Deep tiers which allow for more searches.
Is the Consensus plugin for ChatGPT free?
Consensus for ChatGPT is free. It delivers more or less the same functionality when searching on the Consensus website (which is free too). However, to get insights from within individual papers you will need to use the Consensus website and their AI to analyse the results in-depth.
Summary
Consensus is a powerful AI tool that helps you discover papers and jump into a literature review. It truly shines when you ask yes/no questions because its interface has been designed with these kinds of questions in mind. It is equally useful in an academic context as it is for personal research and day-to-day queries.
To recap, here’s how Consensus works:
- Consensus starts with a plain text query: your research questions or keywords.
- Consensus answers your query with a summary and corresponding papers. If you use Deep Search, you’ll get an extremely thorough analysis and essentially an entire mini lit review, based on 50 papers.
- Ask it yes/no question: Consensus AI analyses how each paper leans and displays that as a small yes/no/possibly visual, showing you the real “consensus” of the literature.
- Consensus is available as a custom GPT in ChatGPT, which allows for unique use cases as you can leverage ChatGPT’s conversational capabilities and Consensus’ database of studies.
- Drawbacks to be aware of are: No deep linking (i.e. you must read a PDF in detail to see why Consensus cited it) and the app itself can get laggy for some tasks or demand a lot of CPU from your browser.
- You can use Consensus AI for free, up to a certain limit. For the premium tier, the price is $10/month. Get 30% with my discount code EFFORTLESS30.
Consensus isn’t just a great tool all by itself, but integrates well into a broader literature review workflow.
👉 Check out the AI Agents for Academics course if you want to learn how to build AI-powered workflows systematically using Obsidian + the latest plugins and frontier AI integrations.


