Consensus searches over 200 million peer-reviewed papers and answers claim-shaped questions — does intermittent fasting improve metabolic markers, does this drug work — showing what proportion of studies support, contradict or are neutral on the claim. That consensus meter is the product, and it's genuinely useful for getting oriented fast. It's also the thing to hold lightly: a percentage of papers is not a weighting by study quality, and ten small studies don't outrank one large trial no matter what the bar chart shows. Paid plans run around $10 to $15 a month with a usable free tier.
Reviewed by EdGrows·Updated Aug 18, 2026Affiliate link — we may earn a commission
This analysis is based on documentation, public user reports, and vendor materials — not yet on our own hands-on testing. How we rate
Most research tools help you find papers and leave the interpreting to you. Consensus does something narrower and more immediately useful: it tells you whether the literature agrees.
Ask a claim-shaped question — does creatine improve strength, is intermittent fasting effective for weight loss, does this intervention reduce readmissions — and it searches over 200 million peer-reviewed papers and returns a distribution: what proportion of studies support the claim, contradict it, or sit neutral.
That's the consensus meter, and it answers the question most people actually have. Not "what papers exist about this" but "is this thing real."
For journalists on deadline, clinicians orienting quickly, students, and anyone who wants a sourced read before forming an opinion, it's the fastest useful tool in this category.
The meter counts papers. It does not weight them.
That means:
None of that makes it useless — it makes it a signal rather than a verdict. The right use is seeing whether a question is settled, contested or thinly studied, then reading the actual papers before concluding anything that matters.
A proper meta-analysis weights by sample size, design and risk of bias, and takes months. A consensus meter takes four seconds. Those are different instruments and confusing them is the main way this tool causes harm.
Worth being direct about one application: this isn't for medical decisions. It's useful for understanding a topic before a conversation with a clinician, and it doesn't replace one.
Reported pricing runs a free tier, paid plans around $9.99 to $14.99 monthly, with one 2026 comparison describing Consensus as a $144-a-year starting point — roughly $12 on annual billing. Enterprise is quoted separately.
The price is a substantial part of the appeal. Elicit's Pro tier reportedly runs around $588 a year; Consensus does its narrower job for a fraction of that, and the free tier is genuinely usable rather than a demo.
That combination is why it keeps getting recommended as the starting point in this category. Most people asking research questions don't need systematic review tooling — they need a fast, sourced answer, and paying review-tool prices for that is overspending.
Works well: empirical claims with a body of published research behind them. Does X affect Y. Is there evidence for Z. Do studies agree.
Works badly: open theoretical questions, humanities scholarship, conceptual debates, anything where the interesting work isn't testable. There's nothing for a meter to measure, and the interface will produce something anyway — which is worth knowing before trusting it.
Coverage skews toward well-indexed empirical fields, as it does across this whole category. Niche subfields and non-English literature are thinner regardless of how polished the tool is.
Serious research uses several tools at different stages, and Consensus occupies one clear slot:
Is this claim supported? → Consensus What does this field look like? → ResearchRabbit Extract data across 40 papers → Elicit Exhaustive answer to one hard question → Undermind Synthesise my own document pack → NotebookLM
Nobody needs all five. The useful framing is which stage you're stuck at, not which tool ranks highest — and for a great many people, the stage they're stuck at is "is this true," which is exactly what Consensus answers.
Weighted evidence assessment. A percentage isn't a meta-analysis.
Non-empirical fields. No claims to measure.
Systematic review workflow. No extraction, screening or export — that's Elicit.
Clinical decisions. Orientation, not advice.
Against Elicit: Elicit does screening, extraction and export for formal reviews at a reported $588 a year on Pro. Consensus answers a claim in seconds for around $12 a month. The 2026 comparisons that call Elicit best for systematic reviews and Consensus the better starting point are getting it right — different jobs, and most people need the cheaper one.
Against ResearchRabbit: ResearchRabbit maps citation networks for discovery and now has a ~$10 RR+ tier alongside its free plan. Consensus answers questions; ResearchRabbit shows you a field. Complementary.
Against Undermind: Undermind runs deep agentic searches and returns a written synthesis on one question. Slower, deeper, and better when the question is genuinely hard rather than genuinely common.
Against Perplexity: Perplexity searches the live web with citations, which is right when the evidence isn't in journals — policy, news, current events. Consensus is right when it is.
Against ChatGPT: ChatGPT will answer confidently and may invent supporting citations. For any claim you'll repeat publicly, that's the reason Consensus exists.
| Plan | Reported | For |
|---|---|---|
| Free | $0 | Occasional questions, genuinely usable |
| Premium | ~$9.99–12/mo | Regular use, higher limits |
| Pro | ~$14.99/mo | Heavier use, additional features |
| Enterprise | Custom | Institutions and teams |
Checked August 2026. Reported figures vary slightly across sources — paid tiers appear between $9.99 and $14.99, with one 2026 comparison citing $144 a year as the effective starting point on annual billing. Corpus reported at over 200 million papers. Confirm current plans on consensus.app before subscribing.
Read the meter as a shape, not a score. Settled, contested or thin — that's the information.
Click through to the papers when it matters. Especially if the answer is one you'll repeat in public.
Start free. It's usable enough that many people never need to upgrade.
Don't use it as a substitute for a clinician. Orientation before the conversation, not instead of it.
Consensus does one thing exceptionally well: it tells you, in seconds and with sources, whether the published literature supports a claim. For journalists, students, clinicians orienting quickly and anyone who wants a grounded read before forming a view, that's more immediately useful than a list of search results, and at a reported $10 to $15 a month with a genuinely usable free tier it's the most accessible entry point in academic AI search.
The limitation is structural and worth repeating whenever the meter looks decisive. It counts papers rather than weighting them — ten small observational studies can visually outweigh one large trial, publication bias is inherited wholesale, and study quality is invisible in a percentage. That makes it a signal about the shape of the evidence, not a verdict on the question, and the distinction matters most exactly when the answer looks clearest.
Two narrower limits: non-empirical fields don't fit the model at all, and coverage skews toward well-indexed empirical literature as it does across the category.
For anyone who needs to know what research says about a claim — which is most people, most of the time — recommend, starting free. For formal systematic reviews, Elicit is the tool and the price difference is justified. For medical decisions, neither replaces a clinician.
Note: Consensus does not currently have an affiliate program with AIVario. We earn no commission, and this rating carries no commercial incentive.
Best for: Fast sourced answers to empirical claims, journalists and writers checking assertions, students orienting in a topic, anyone wanting evidence before an opinion Not ideal for: Weighted evidence assessment, humanities and theoretical questions, systematic review workflow, clinical decision-making Bottom line: The fastest way to see whether research supports a claim, at a price that makes it the sensible starting point — as long as you read the meter as a shape rather than a score.
There's a free tier, with paid plans reported around $9.99 to $14.99 per month depending on tier, and one 2026 comparison describing it as a $144-a-year starting point — roughly $12 monthly on annual billing. Enterprise is quoted separately. Figures vary a little across sources, so confirm directly. Relative to Elicit's Pro tier at a reported $588 a year, Consensus sits firmly in the accessible bracket, which is much of its appeal.
A visual summary showing what proportion of retrieved studies support, contradict or take a neutral position on your claim. Ask whether creatine improves strength and you get a distribution rather than a single answer, with the underlying papers listed. It's the fastest way to see whether a question is settled, contested or thinly studied — which is often the most useful thing to know before reading anything in depth.
As a signal, yes. As a verdict, no, and this matters. The meter counts papers rather than weighting them by sample size, study design or quality, so ten small observational studies can visually outweigh one large randomised trial. It also reflects what's been published, which carries publication bias toward positive findings. Use it to see the shape of the evidence, then read the actual papers before concluding anything that matters.
Different jobs at very different prices. Consensus answers a claim-shaped question quickly and cheaply. Elicit does the heavy work of a systematic review — screening, structured extraction, comparison, export — at a reported $588 a year for Pro. One 2026 comparison called Consensus the better starting point for most people and Elicit the right tool for a formal review, which is a fair summary of both.
No, and no tool in this category should be used that way. Consensus surfaces what the literature contains, not what applies to an individual case, and it does not weight evidence quality or account for clinical context. It's genuinely useful for understanding a topic before a conversation with a clinician, and it is not a substitute for one. That distinction is worth keeping firm even when the meter looks decisive.
Questions that are shaped like claims and have been studied empirically. Does X improve Y, is there evidence for Z, do studies agree about this mechanism. It's fast, sourced, and answers the question most people actually have — is this thing real — without requiring them to read twelve abstracts first. Journalists, clinicians orienting quickly, students and curious non-specialists get the most from it.
Anything not shaped like a testable claim. Open questions, theoretical debates, humanities scholarship and topics where the interesting work is conceptual rather than empirical don't fit the model — there's nothing for a meter to measure. Coverage also skews toward well-indexed empirical fields, so a niche or non-English literature may be thinly represented regardless of how good the interface is.
For occasional questions, yes — it's genuinely usable rather than a demo, which is why Consensus gets recommended as a starting point so often. Paid tiers raise limits and add features for people using it as part of regular work. The honest test is the same as everywhere: upgrade when a limit interrupts you, not because the paid feature list looks appealing.
The best sourced-answer engine there is, and the browser is now free. Pro at $20 is the tier; Max at $200 is a different product.
Still the best free AI assistant available, with no paid tier to upsell you. API pricing got more complicated in August.
Google's source-grounded research tool. The free tier limits how much you use, not what you can do.