What the measurement actually costs (our own bill)
Every tool in this category pays the same underlying costs: API calls to the engines it measures. Two vendors deserve credit for partial disclosure: Rankscale publishes its credit mechanics openly (most engine queries cost about 0.25 credits per prompt, with a live calculator), and Sight AI publishes full credit conversions (an all-platform visibility check runs about 31 credits). What nobody else in the category publishes is the reconciled dollar bill. On 2026-08-06 we ran our full weekly panel and reconciled the ledger against the invoice: 147 ChatGPT probes with web search enabled billed $51, about $0.34 per probe. The web-search tool fees and retrieved-page context, not the base tokens, are most of that cost.
A panel day for one site on our instrument, one query set across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews with repeat sampling on high-intent queries, costs us somewhere in the range of $20 to $80 in engine fees. The spread is real and worth stating rather than averaging away: the bill scales with how many queries the set carries and how many times each is re-sampled, and ChatGPT with web search dominates it. Note also what that figure does not include. It is one query set for one site. A site measured on several panels pays that several times over, and benchmarking a set of competitors alongside your own brand multiplies the probe count again, which is why competitor tracking is the part of this category that gets expensive fastest. Weekly cadence puts the raw measurement floor for one seriously-measured site in the low hundreds of dollars per month before any analysis, storage, or human time. When a vendor charges $99 a month, this is what the margin is absorbing; when a tool is much cheaper, it is usually sampling less, covering fewer engines, or running without web search, which our own testing showed measurably changes the answers you get.
Figures from our 2026-08-06 billing reconciliation, published in full in our research record. Engine API prices change; this is a dated reading.
What the tools charge
On 2026-08-07 we captured the live pricing pages of eight vendors in this category directly, rather than relying on third-party roundups. The published floor is $29 per month (Otterly.AI Lite: 15 prompts, 4 engines), with HubSpot's beta AEO product at $50. Four of the eight anchor an entry or near-entry tier at $99: Profound, SE Visible, Sight AI, and Rankscale. Above the anchor, published self-serve pricing runs higher than the category's reputation suggests: $399 (Profound Growth), $489 (Otterly Premium), $500 month-to-month (Scrunch Growth), $780 (Rankscale Enterprise), and Sight AI's enterprise tier starts at $2,500. One caution from the capture: engine counts are not comparable at face value. Otterly's base price covers four engines with Claude, Gemini, and Google AI Mode as paid add-ons; Profound's $99 tier tracks ChatGPT only; Scrunch and Rankscale include their full engine set in every tier. A price-per-engine comparison that ignores the bundling rules will mislead you.
Our own pricing, since a cost page that hides its author's prices fails its own test: $99 per month for measurement only (two panel runs a month), $399 per month for weekly measurement plus four content drafts, and $1,500 per month for Compete, which is weekly measurement plus two drafts a week. Not every client gets exclusivity. Metro grain is available soon; today exclusivity is offered at state, national, or online grain. Our Operated engagements are quoted per engagement rather than list-priced, so they carry no published price.
Category figures captured 2026-08-07 directly from eight vendors' live pricing pages (primary observation; capture record in our research archive). Prices change — verify with each vendor. Our prices are current and binding as published on our pricing page.
The free option, honestly
You can measure the direction of your own AI visibility in an afternoon for nothing. Ask each major assistant about your business by name, then ask the category questions your buyers actually ask without your name in them, in fresh chats, and record which brands come back. Done monthly, that tells you whether you're invisible, present, or slipping on the handful of questions that matter most.
What the manual version cannot do is what you'd be paying for: repeat sampling (single AI answers vary run to run; our instrument asks high-intent questions three times per engine and takes the majority), a frozen question panel so week-over-week movement is real rather than rephrasing noise, competitor rates measured on the same questions, and the paper trail of which pages the engines actually cited. If the manual check says you have a problem worth working on, that is when paying for instrumentation starts to make sense, ours or anyone's.
How to compare quotes
Four questions expose most of the difference between tools at the same price: How many engines, are they all in the base price or paid add-ons, and is web search on when they probe? How often does it re-measure, and is the question set frozen between runs? Does it show you the underlying answers and cited pages, or only a score? And does the vendor publish its methodology and its own numbers anywhere you can check? The last one is rarer than it should be: in our audit of the ten most-cited comparison pages in this category, about one in four quantitative claims carried any source at all.
Common questions
Why do AI visibility tools cost more than rank trackers?+
Because every data point is a paid AI answer, not a scraped results page. A web-search ChatGPT probe billed us about $0.34; multiply by engines, queries, repeat samples, and weeks, and the raw engine fees for one site can exceed a rank tracker's entire subscription. Tools priced far below that floor are measuring less than it appears.
Is there a genuinely free AI visibility tool?+
Free tiers exist and are useful for a first look, and the manual check costs nothing. Both are one-time snapshots without repeat sampling or a frozen panel, so they can tell you where you stand today but not whether anything you do is working. That distinction is what the paid tier is actually buying.
What does your own measurement cost you?+
About $0.34 per web-search ChatGPT probe, per our 2026-08-06 billing reconciliation, which works out to somewhere in the range of $20 to $80 in total engine fees for a five-engine panel day on one query set for one site. Measuring several query sets, or benchmarking competitors alongside your own brand, multiplies that many times over. We publish this because cost transparency is cheap for us and expensive for anyone whose pricing depends on the measurement sounding mysterious.
keep reading
See what AI says about your brand right now — free, on a live engine.
Scan your domain