Semrush AI Visibility Toolkit pricing and what it includes

The first number on Semrush’s AI pricing page is £80.22 a month per domain, billed annually. That is the UK price, checked on 2 Oct 2026. The page pairs that price with AI visibility reports, competitor research, prompt research and tracking across ChatGPT, Google AI, Gemini and Perplexity. It also lists additional users as an add-on, so the displayed amount is only one part of the buying decision.

Add-on price

Semrush’s pricing page says “£80.22/mo per domainbilled annually”. It also lists “1 domain for Brand Performance analysis”, which makes the domain the visible billing unit. Additional users start at £36 per month, according to the same page, while unlimited projects, custom limits and custom large-scale AI prompt tracking are listed without fixed prices.

The page does not give us a base-plan price or state whether a separate core Semrush subscription is required. We can therefore verify the AI Toolkit charge, but not the complete monthly bill for a new customer. The £36 user add-on would also change the comparison quickly for a team, so record it separately rather than treating £80.22 as the final figure. When we switched the page to monthly billing on 2 Oct 2026, it showed the same £80.22 a month per domain, with no billing note.

The same page appears to list a daily allowance of AI analysis reports, beside AI Visibility Overview, Competitor Research and Prompt Research. Its formatting does not make the relevant tier clear, so we are not using that row as evidence for a clean entitlement until Semrush confirms it.

Toolkit scope

Semrush describes the product as being for “businesses that want to track how they’re mentioned in AI search”. Its listed scope includes “AI visibility reports for any domain”, “AI competitor analysis and prompt research” and “Site audit for AI readiness”. These cover monitoring mentions, comparing competitors, researching prompts and checking site readiness. Treating them as one score can hide which job you are paying to perform.

The monitoring layer includes mentions from ChatGPT, Google AI, Gemini and Perplexity, according to the pricing page. It also lists daily, weekly and monthly data updates. That cadence helps when you need to watch movement over time instead of judging the product from one answer. It does not show whether every question, answer or source page receives the same treatment.

The prompt layer is more specific. Semrush says it can “Track 25 custom prompts (daily AI rankings)”. That gives you a defined recurring prompt allowance, not full coverage of every way a buyer could describe a category, use case or comparison. The page also lists prompt research, but it does not explain how that research works or how many prompts it can discover.

The domain wording needs care. Brand Performance analysis is listed for one domain, while the wider description refers to reports for any domain. We can report both statements, but the public page does not explain whether they describe separate screens, limits or modes. That ambiguity would need to be confirmed directly with Semrush before it goes in a comparison.

Visibility score

A visibility score is a roll-up. It turns observations about mentions, rankings or brand performance into a summary that helps you compare a domain with itself, competitors or an earlier period. Semrush’s public copy confirms AI visibility reporting and Brand Performance analysis, but it does not publish the score formula, the weighting of engines or the treatment of citations. We therefore cannot infer what a particular score means without seeing the report behind it.

A prompt-level scan answers a narrower question: what did an assistant return for selected prompts, on selected engines, in selected runs, at a particular point in time. CiteGraph Starter makes that unit explicit with two projects, ten questions per project, three runs per question, four engines and six tracked rivals. Those details describe prompt evidence rather than a general statement about a domain’s overall visibility. A prompt scan keeps the run count and engine visible, so you can check the claim directly instead of trusting a summary number.

The distinction matters when a score moves. A higher score could reflect changes across the tracked data, but the score alone may not show which question changed, which competitor appeared or which page was cited. A prompt scan gives you a more inspectable trail, provided you preserve the prompt, engine and run details. Neither method makes the other redundant.

There is a fair case against our position: a roll-up is more useful than a pile of individual answers when a marketing team needs a recurring report. That is true. A score gives you a compact way to spot direction and decide where to look next. The answer is to use it as an index, then open prompt-level evidence before changing a page, campaign or claim.

Prompt coverage

The prompt allowance is the clearest constraint in the Semrush AI Visibility Toolkit pricing material. Semrush lists 25 custom prompts with daily AI rankings. That gives you a starting set for recurring checks, but it is not the same as covering every buyer question in a market. If your list comes from internal assumptions, it can miss the question that matters most, so prompt research and prompt tracking should be treated as separate activities even when one product offers both.

Use this checklist before comparing the Toolkit with a prompt-level product:

• Write down the exact domains you need to analyse, because the published price is stated per domain.
• Separate recurring prompts from one-off research questions, because daily rankings and exploratory research serve different jobs.
• Confirm which engines your buyers actually use before paying for coverage of all four named engines.
• Ask how the visibility score is calculated, because the public pricing copy does not publish its formula.
• Ask whether the report shows the answer and cited page behind each change, rather than only the aggregate result.
• Check whether the report limits match the number of prompts and domains your team plans to track.
• Add selected user or usage add-ons to the comparison, because the published AI price is not the complete bill, as noted above.

For a first check of the underlying site, our free AI readiness check needs no account and shows the fix for each of its fifteen checks. It does not replace visibility monitoring or competitor research. It answers a different question: whether common site signals are ready to be read and used by AI systems. A focused prompt set can still be enough for brand monitoring, provided you know what it leaves out.

If you want to inspect answer-level visibility before buying a recurring platform, the free AI visibility checker is a direct starting point. CiteGraph’s paid plans then use projects, questions, runs and tracked rivals as the billing shape. That makes the comparison less elegant than a single score, but more explicit about what gets checked.

Buying decision

Semrush makes sense when you need domain-level reporting, recurring AI rankings, competitor analysis and prompt research in the same product. The published package also includes an AI readiness audit and several update intervals. Those features can reduce the number of separate reports your team maintains, although the public page does not tell us how deeply each report can be exported or customised. Compare the listed AI charge with the users, limits and domains your team actually needs.

A prompt-level scan makes more sense when the immediate job is diagnosis. You want to know whether a buyer question produced a mention, which rival appeared, what sources the answer used and whether the result held across repeated runs. A scan will not necessarily provide the same trend view as a visibility score, but it makes the observed answer easier to inspect. That is the better fit when a page needs a reason, not merely a movement line.

The choice is not Semrush versus Google Analytics. Analytics measures visits and on-site behaviour, while the Semrush AI product is described around mentions, rankings, prompts, competitors and AI readiness. Those are different measurement layers. Adding one does not turn it into the other.

Our data has a firm limit: we measure samples and the product details available in the cited pages, not every Semrush workflow or every assistant answer. We print the boundaries of the comparison, and we do not know what we have not measured. The responsible choice is therefore to compare the published add-on scope with the exact prompts, engines and evidence your team needs.

The Semrush AI Visibility Toolkit is priced per domain at £80.22 a month billed annually, and that figure is not the same thing as a prompt-level scan.

We also compare the toolkit directly with three dedicated trackers: Profound vs Semrush AI Visibility Toolkit, Peec AI vs Semrush, and Ahrefs Brand Radar vs Semrush AI Visibility Toolkit.