Consensus Pricing 2026: Free, Pro, Deep & API Costs

Consensus pricing is unusually clear for an AI research platform. There is a permanent Free tier, Pro costs $20 per month or $144 per year, and Deep costs $65 per month or $540 per year. Team and Enterprise use custom pricing. The major difference between tiers is not access to basic scientific search; it is the amount of advanced synthesis, Study Snapshot and API/MCP capacity.

This makes plan selection relatively straightforward. Researchers who mostly search papers can stay Free for a long time. Users who need frequent cited synthesis and unlimited Pro messages can move to Pro. Deep is for researchers whose workload includes repeated literature reviews or high-volume scientific access through integrations.

Consensus pricing at a glance

Consensus Free: $0

The Free plan provides unlimited basic Papers searches. That matters because users can evaluate the core scholarly-search workflow without being forced into a short trial window. Advanced features are capped at 10 Pro messages, 3 Deep Reviews, 10 Study Snapshots and 30 API/MCP calls per month.

Free is a realistic long-term tier for students, journalists or occasional researchers who mainly need paper discovery. It becomes restrictive when users rely on synthesis rather than search or when Consensus is connected to another AI assistant through MCP.

Consensus Pro: $20/month or $144/year

Pro is the mainstream paid plan. Monthly billing is $20. Annual billing is $144 upfront, equivalent to $12 per month. That is a 40% effective discount compared with twelve monthly payments totaling $240.

Pro removes the monthly cap on standard Pro messages and Study Snapshots. It provides 15 Deep Reviews per month and 500 shared API/MCP calls. For most individual researchers who use Consensus several times per week, this is the natural paid tier.

Who should choose Pro?

Choose Pro if Free's 10 Pro messages or 3 Deep Reviews are regularly constraining your research but 15 Deep Reviews cover a normal month. It is also the most economical route for researchers who want to connect Consensus to ChatGPT, Claude or custom workflows without needing thousands of calls.

The annual plan is compelling for users who already know Consensus will remain part of their workflow. Paying $144 rather than $240 over twelve months creates a large discount, but annual buyers should understand the refund window before committing.

Consensus Deep: $65/month or $540/year

Deep is designed for high-frequency literature-review work. Monthly billing is $65. Annual billing is $540, equivalent to $45 per month and saving $240 versus twelve monthly payments.

The key jump is Deep Reviews: 200 per month instead of 15 on Pro. API/MCP capacity also increases from 500 to 2,000 calls. Researchers who run several deep evidence syntheses every week can therefore avoid constantly managing the Pro ceiling.

Deep is difficult to justify for users who mostly run basic paper searches. The plan is a capacity purchase. A researcher should upgrade when Deep Review or integration usage is consistently limiting output, not because the higher tier sounds more capable.

Team pricing

Consensus Team pricing varies by team size and is not publicly listed as one numerical rate. Current plan documentation places Team between individual and Enterprise usage, with collaborative administration and per-user allowances.

Because the price is custom, teams should request a written quote covering seat minimums, Deep Review limits, API/MCP calls per user, administrative controls, onboarding and renewal terms. A research lab with five users and a department with 100 users can have very different requirements.

Enterprise pricing

Enterprise is also custom. The current SearchSagar research record notes that Enterprise is positioned for organizations with more than 200 users, although contract scope should be confirmed directly with Consensus because institutional deployments can vary.

Enterprise buyers should ask about authentication, organization controls, data handling, API capacity, support, procurement terms and whether additional usage is billed at the same public API overage rates.

API and MCP allowances

Consensus treats API and MCP calls as one shared allowance. Free includes 30 calls, Pro and Team list 500, Deep includes 2,000, and the live API page also lists 500 included calls for Enterprise. This allowance is relevant even for non-developers because MCP integrations with ChatGPT, Claude and Copilot consume the same pool.

Usage beyond the included allowance currently costs $0.05 per API call plus one additional credit for each 100 papers processed. A simple single-search integration will consume far less than a workflow that repeatedly processes large paper sets.

Annual billing savings

The annual discount is one of Consensus's clearest pricing advantages. Pro falls from an effective $20 monthly to $12, while Deep falls from $65 to $45. Researchers with predictable year-round usage can therefore reduce cost materially without changing features.

The tradeoff is commitment. Refund windows for annual plans are measured in days rather than months. Users who are unsure about fit should start Free or use one month of Pro before prepaying a year.

Refund policy

Consensus documents specific refund-request windows. Monthly subscription purchases can request a refund within 72 hours. Annual subscription purchases have a five-day window. Eligible EU customers can receive a refund if they cancel within 14 days of purchase according to current policy.

These are request windows, not a reason to delay evaluation. Users should confirm plan fit quickly after purchase and keep a copy of the applicable billing terms.

Cancellation

Subscriptions can be canceled from Profile > Settings > Subscription. Paid features remain available through the current billing period, after which the account returns to Free. The existence of a useful Free tier makes cancellation less disruptive because saved research can remain associated with the account even after premium capacity ends, subject to current account policies.

Extra costs

• API overage: $0.05 per call beyond the included allowance.
• Paper-processing credits: one additional credit per 100 papers processed under the current API model.
• Team: custom price based on organization size.
• Enterprise: custom institutional terms.
• External tools: Microsoft 365 Copilot or other AI assistants can require their own subscriptions independently from Consensus.

How many Deep Reviews do you really need?

The difference between 15 and 200 Deep Reviews is so large that it should drive the Pro-versus-Deep decision. A researcher running one or two deeper syntheses each week may remain comfortably inside Pro. A team member performing several reviews per working day can exhaust Pro quickly.

Track usage for one month before upgrading. If Pro is constrained only near the end of an unusually busy month, the $45–$65 Deep cost may not be necessary. If the cap is reached every week, Deep is much easier to justify.

Pro versus Deep: a practical break-even framework

Pro costs $20 monthly and Deep costs $65, so the monthly step-up is $45. On annual billing, Pro is $144 and Deep is $540, a $396 yearly difference. That premium is mainly buying far more Deep Review capacity and four times the API/MCP allowance.

If a researcher consistently uses more than 15 Deep Reviews per month, Deep becomes easier to justify. If usage fluctuates around five to ten reviews, Pro is financially stronger. The same logic applies to integrations: users approaching 500 MCP or API calls every month should compare Deep with paying overage at $0.05 per additional call.

For example, 1,500 extra API calls beyond Pro's included allowance would cost about $75 before additional paper-processing credits. At that level, the $45 monthly difference to Deep may be cheaper while also providing dramatically more Deep Review capacity. Lower overage volumes favor staying on Pro.

How Team and Enterprise buyers should compare quotes

Custom pricing makes organizational comparison harder. A quote should specify seat count, whether research allowances are allocated per user or pooled, included API/MCP calls, Deep Review capacity, onboarding, support level, renewal increases and any minimum commitment.

Institutions should also ask whether student, researcher and administrator roles can have different access. A university may have thousands of occasional users but only a smaller group performing intensive research. A contract that treats every user as a high-capacity seat could be unnecessarily expensive.

External subscription costs

Consensus integrations can create value without replacing the host platform. Using Consensus inside ChatGPT, Claude or Microsoft 365 Copilot may require a paid subscription to those products in addition to Consensus. Zotero itself has its own storage economics for users who exceed free cloud capacity.

Organizations should therefore calculate the cost of the complete research stack. Consensus may be inexpensive at $12 annual-equivalent on Pro, but an institutional workflow involving Microsoft 365 Copilot, reference management and other research databases can have much higher total software cost.

Budgeting API and MCP automation

Interactive use is easy to control because a human initiates each search. Automation is different. An agent can generate multiple searches, retrieve large result sets and repeat steps without the user seeing every API call. This can consume the included allowance much faster than manual use.

Developers should log requests, cache repeated queries and set application-level limits. If the workflow runs on a schedule, calculate the maximum theoretical call volume before deployment. At the current $0.05 overage rate, small inefficiencies become visible at thousands of calls.

Monthly versus annual billing

Annual billing is financially attractive but should follow a real usage test. Pro annual saves $96 compared with twelve monthly payments, while Deep annual saves $240. Those are large discounts, but the annual refund-request window is only five days under current policy.

A cautious buyer can remain on Free, purchase one month of Pro, and measure research behavior before switching to annual. Deep buyers should similarly confirm that high review volume is persistent rather than tied to one temporary project.

Cost scenarios

A graduate student who runs basic searches daily and only a few advanced reviews can stay Free. A clinician or policy researcher who asks synthesis questions throughout the week will usually fit Pro. A systematic-review team member or evidence analyst running large numbers of deep reviews may benefit from Deep.

For developers embedding Consensus in a product, API consumption can matter more than the subscription price. At $0.05 per extra call, 10,000 overage calls would add $500 before paper-processing credits. Automated workflows should therefore have explicit call limits and caching where appropriate.

Consensus pricing versus alternatives

Elicit offers a Free tier and published annual individual rates starting at $11 per user per month for Plus, with higher Pro and Scale tiers for systematic reviews and collaboration. SciSpace has a Free Basic tier with 100 Agent credits, Premium at $20 monthly or $12 annual-equivalent, Advanced at $90 monthly or $70 annual-equivalent and Max at $200 monthly or $160 annual-equivalent. Scite's current annual-effective pricing lists Basic at $20 per month and Pro at $50.

Consensus is especially competitive at the Pro annual rate of $12 per month. Elicit becomes more attractive for rigorous systematic-review screening and extraction, SciSpace for an end-to-end research/writing workspace, and Scite for citation-context analysis.

Final pricing verdict

Consensus has a well-designed pricing ladder. Free is useful enough for real research, Pro is inexpensive on annual billing, and Deep charges a clear premium for users who genuinely need far more literature-review capacity.

The biggest uncertainty begins above the individual tiers because Team and Enterprise are custom. For individuals, however, plan economics are more transparent than many AI research products. Most buyers can choose accurately by measuring how often they need Deep Reviews and connected-tool calls.

Frequently asked questions