Research

Research reportgeo

Subscription or Nothing: How 72 AI-Visibility Tools Bill

AI-visibility cost scales with usage. Billing doesn't: of the 69 tools where we could determine the model, 59 sell a subscription and one bills per unit consumed.

The work an AI-visibility tool does for you is usage-shaped. Its cost is driven by how many prompts you track, across how many engines, refreshed how often — and every one of those is a dial you might want to turn up for a product launch and back down again in a quiet quarter. So it is worth knowing, before you shortlist anything, whether this category will actually sell you that flexibility. We went through all 72 AI-visibility and GEO software tools in our index and recorded how each one takes money.

Of the 69 tools where we could determine a billing model, 49 sell a flat recurring subscription and 10 more are freemium with a subscription above the free tier: 59 of 69 between them. 5 are free. 5 meter by usage, and four of those five are enterprise platforms that publish no rate card at all, which leaves one tool in the index billing purely for what you consume. The remaining 3 of 72 we could not determine, which is not the same as a finding.

How we counted

Listings in our index carry a pricingModel field where we have been able to establish one: a validated enum with five permitted values (free, freemium, subscription, usage-based, one-time), researched against the vendor’s own live pricing page and shown publicly on each listing’s spec sheet. This census is a straight count of that field across the corpus on 2026-08-13.

The denominator is the 72 software tools, not all 87 published listings. The 15 managed citation-services agencies sell a team rather than a plan, so “how does it bill” is a different question for them, and they are excluded here, consistent with how this site’s other census articles scope the same corpus. Every one of the 72 was last verified against its own pricing material between 2026-07-23 and 2026-08-12.

Three tools carry no recorded billing model, so they sit outside every figure below: Addlly AI, Big Leads and Webglazer. They are unknown to us, not confirmed as anything, and this is why the counts above are quoted out of 69 rather than out of 72.

One thing this census is not measuring: whether a tool has a free tier. That is a separate axis with its own free-tier census, which partitions the corpus by the best free option a buyer can get. A tool billed subscription here can still run a free tier, and several do. The two sets of numbers answer different questions and should not be added together.

What the 69 look like

Billing model Count Share of the 69 What it means for a buyer
Subscription 49 71% A recurring monthly or annual plan. Usage limits are attached to the tier you buy, not billed separately.
Freemium 10 14% A standing free tier with recurring paid plans above it. Paid access is still a subscription.
Free 5 7% No paid tier in our record at all, including open-source and bring-your-own-API-key tools.
Usage-based 5 7% Cost tracks consumption rather than a fixed period — but see below, because this bucket is not one thing.
One-time 0 0% No tool in the index is recorded as selling a perpetual licence as its primary model.

A caution on that last row, because a zero is the easiest number on this page to misread. It means no tool’s primary model is a one-time purchase. It does not mean no one-time purchase exists anywhere in the category: visibai, for example, offers a one-time €29 starter audit alongside its monthly plans and is recorded here as subscription, which is what its main ladder is. Read that row as “nobody sells this category as a perpetual licence,” not as “no one-off option is purchasable.”

The usage-based five, which are not five of the same thing

This is the bucket that matters most for the question at the top, and it is the one where the count alone would mislead you. Four of the five are large enterprise platforms where AI-visibility is a component of a much bigger contract, and none of the four publishes a rate card:

Tool What its own pricing material publishes
Adobe Brand Visibility No standalone price. Sold inside a custom Adobe CX Enterprise contract; no self-serve signup.
Cision AI Visibility Dashboard No published list price. Bundled into CisionOne, which is sales-gated.
Conductor No public rate card. Three named tiers (Essentials, Growth, Enterprise), each quoted as a custom annual contract.
Meltwater GenAI Lens No public price. A prorated add-on that requires an existing Meltwater subscription.
Sleepwalker €0.02 per credit, €5 minimum top-up, no monthly plan. A single-engine visibility check costs 1 credit; reading stored results is free.

“Usage-based” on the first four means the quote scales with volume — prompts per year, pages analysed, seats. It does not mean you can pay for a spike and stop. All four are annual contracts, negotiated, with no published entry point. Only Sleepwalker meters in the way a buyer reading “usage-based” would expect: top up €5, spend it, stop.

So the honest version of the headline is that one tool of the 69 lets you pay for what you use without either a subscription or a negotiated contract. That is the number we would want quoted, and it is a considerably smaller claim than “5 of 69 are usage-based.”

Why this is the shape of the market

Two forces point the same way, and neither is a vendor being difficult.

The first is that the vendor’s own cost is usage-shaped too. Running a tracked prompt means paying an LLM provider per call, and a tool tracking 400 prompts daily across six engines is spending real money on the buyer’s behalf every day. A flat subscription with a hard prompt cap is the simplest way to make that cost predictable on the vendor’s side. This is why the prompt-volume ladder is the most common shape in the category, as our walkthrough of pricing units found — the tier is the usage cap, so subscription and usage pricing are less opposed here than the labels suggest.

The second is that recurring revenue is what this category is being funded to show, and a per-credit meter produces worse revenue predictability than a monthly plan.

The consequence for a buyer is concrete. AI-visibility measurement has an obvious low-intensity mode: checking quarterly whether anything moved. The market has priced it almost exclusively for the continuous-monitoring mode. If your actual need is four scans a year, you are mostly choosing between paying twelve months for four months of value, going to the free bucket, or Sleepwalker.

What would change this conclusion

Two things, and both are cheap to check against this page next year.

If the usage-based count rises above 5 while the enterprise share of it falls (that is, if self-serve metered options appear), the category is unbundling and the argument above weakens. If instead the free and freemium buckets (15 of 69 today) shrink as those tools add paid tiers, the commitment problem gets worse, not better.

We would also revise this if the 3 unknowns turn out to break the pattern, though at 3 of 72 they cannot move the headline much in either direction.

Method and reuse

Counts are from the citedindex corpus on 2026-08-13, across the 72 published AI-visibility and GEO software listings (87 published listings less the 15 managed citation-services agencies). Billing model is a validated five-value enum recorded per listing from the vendor’s own published pricing page, verified between 2026-07-23 and 2026-08-12. Denominator for every percentage above is the 69 listings where the field is recorded; the 3 unrecorded listings are excluded rather than counted as any value.

If you cite these figures, please quote them against the 69 (“of the 69 AI-visibility tools where CitedIndex could determine a billing model”) rather than against the category as a whole. We index this market; we have not surveyed it, and the difference matters.

Get the next report

New tools rankings and fresh data reports. One short email, straight to your inbox. One-click unsubscribe.