Authoritas AI Search Pricing: Credits, Plans and Overage

Authoritas AI Search starts at $119/month for 2,000 prompt credits on P1. The free option lists 50 credits; P2 is $299/month for 6,000 and P3 is $499/month for 15,000. These are the official USD selector's prices, checked October 5, 2026, not conversions from its default GBP display. Tax and your contracting region still need confirmation at checkout. Official pricing.

Our view: buy the model access you need before shopping for the cheapest credit. A large allowance is poor value if the selected collection method does not resemble the AI experience your buyers use. This is a documentation-based buyer guide: we operated the public price controls, but did not test a paid Authoritas account or its measurement accuracy.

What does each Authoritas AI Search plan cost?

The self-serve plans charge for credits, with a different overage rate at each tier. Keep monthly and annual billing separate: the annual screen changes both the headline amount and the displayed overage figure. The following table records the monthly USD view, not a monthly equivalent of an annual contract.

Source: Authoritas plan and overage table. Those are AI Search plans. Do not substitute the vendor's separate SEO or API product prices, or assume a different product's data entitlement transfers automatically. Its product selector separates those purchases.

Before approving a subscription, save the selected currency, billing mode, credit allocation and model list together. A screenshot showing only the headline price leaves out the variables most likely to change your bill. Give finance the annual commitment and the expected overage separately.

How do credits translate into usable monitoring?

Calculate usage as sites × questions per site × scheduled runs × the sum of model credit weights. Authoritas lists one-, two- and five-credit model categories; search-grounded options have an Enterprise restriction. The exact model and collection method matter more than the familiar assistant name. Model tiers.

Consider an illustrative monthly schedule: one site, 25 buyer questions, 30 daily runs and two selected one-credit models. That needs 1 × 25 × 30 × (1 + 1) = 1,500 credits. P1 leaves 500 credits of headroom. In a 31-day month, the same design needs 1,550, leaving 450.

Now double the questions to 50. Usage becomes 3,000 credits in the 30-day example. The decision is no longer “does P1 cover 50 prompts?” It is whether P1 plus 1,000 excess credits is cheaper than P2, and whether both plans permit the same selected models.

These are capacity calculations, not measured runs or a promise of exact invoices. Manual reruns, failed-query billing, rounding and special model access need confirmation. Record them in a short audit methodology before treating the resulting series as comparable over time.

When is overage cheaper than upgrading?

At 3,000 monthly credits, P1 plus the published monthly overage is $119 + $66 = $185, below P2's $299. At 5,000 credits, P1 becomes $119 + 3 × $66 = $317, above P2. This comparison assumes proportional overage charging and identical required features; confirm invoicing rules before relying on it.

The approximate cost crossing is 2,000 + ($299 − $119)/$66 × 1,000, or 4,727 credits. That is a budgeting estimate, not an Authoritas plan boundary. A team expecting 4,800 credits should compare P2; a team expecting 3,000 should ask why it needs the larger commitment.

The broader lesson is to use realistic demand, not every possible prompt. A weekly check can be enough for a slowly changing product category. Daily tracking becomes more useful around a launch, a substantial website change, or a competitor event. Increasing frequency improves temporal detail; it does not automatically improve the validity of your visibility measurement.

Keep a stable core of purchase questions and a separate experimental group. Otherwise, a cheaper bill after deleting difficult prompts can masquerade as better brand performance.

Is annual billing worth the commitment?

P1's annual USD screen shows $1,309/year, compared with $1,428 for twelve monthly payments. The arithmetic is $119 × 12 − $1,309 = $119 saved, or one monthly payment. The equivalent is about $109.08 per month, but the annual cash commitment remains $1,309. Annual billing selector.

Do not apply the monthly overage rate blindly to that annual screen. It displays a different overage amount, so obtain a written explanation of the billing period and what one overage unit buys. We have not verified an annual overage invoice and do not model one here.

Start monthly if your prompt set or model requirements are unsettled. An early test may reveal that your buyers care about a different region, that a brand alias creates false matches, or that the response export lacks context you need. A discount cannot compensate for collecting the wrong evidence for a year.

The free 50-credit allocation is suitable for checking a small sample, not proving long-term consistency. With two one-credit model calls per question, 25 questions would consume that allocation once. Confirm its renewal conditions rather than assuming the first displayed allowance guarantees a recurring free research program.

What should you inspect before paying?

Inspect one complete answer and its cited sources, then reconcile it with the dashboard's interpretation. A mention, a recommendation and a linked citation answer different business questions. A score without the underlying evidence is difficult to challenge when your sales team sees something different.

Use this small acceptance exercise; it is a proposed buyer test, not a test we performed:

1. Choose five unbranded purchase questions from real sales conversations.
2. Include one question where your product is unsuitable, to catch overly generous recommendation coding.
3. Save model, collection method, market, timestamp and the full answer.
4. Verify the brand match and each cited URL manually.
5. Export the records and check whether a colleague can reproduce your interpretation.

AEOeye's own audit records, published in the aggregate dataset, contain 892 answers across 127 completed, self-selected audits. Its engine samples are unequal and unsuitable for a vendor accuracy comparison. It illustrates why we keep answers as the unit of evidence rather than treating a dashboard percentage as self-explanatory. See the recommendation measurement guide.

Who should choose Authoritas, and who should wait?

Shortlist Authoritas when you want explicit credit arithmetic and can specify collection methods, cadence and export requirements. Wait if nobody will review the answers or implement changes: a recurring monitoring bill does not create an operating process by itself.

For a suite-level procurement comparison, Conductor's AI Search allowance uses a different published period, while BrightEdge AI Catalyst requires a platform quote. Credits, prompts and bundled access are not interchangeable units. Ask each vendor to price the same workload and return the same evidence sample.

If you only need a baseline before selecting recurring software, AEOeye's free audit is the simpler starting point. Its stated scope is three buyer questions on ChatGPT; the full report is $29 once, rather than a monitoring subscription. It cannot replace ongoing Authoritas tracking, and neither option guarantees that an assistant will recommend you.

Our purchase rule is straightforward: choose the smallest plan that supports a useful, reviewable study, then expand when the resulting decisions justify more coverage.