AI visibility tracking tools: how they work and what they cost
AI visibility tools send prompts to assistants such as ChatGPT and Gemini, store the answers and count which brands and sources appear. Public prices we checked on 4 October 2026 run from $29 to several hundred dollars a month, and the price mostly buys prompts, engines and runs.
In short
- An AI visibility tool sends prompts to AI assistants on a schedule, stores each answer and counts the brands named and the sources linked.
- There are two designs, a prompt panel that you define and a large prebuilt prompt database that you search like a keyword index.
- On 4 October 2026, vendor pages listed entry plans at $29 a month for 15 prompts (Otterly) and $99 a month for 200 prompts (SE Visible), so price per prompt varies widely.
- Plans are priced by number of prompts, number of assistants and how often prompts run, and extra assistants are often paid add-ons.
- Search Console, Bing Webmaster Tools and GA4 report part of the picture at no cost, and none of them shows what ChatGPT says about your brand.
An AI visibility tool does one thing at scale: it sends prompts to AI assistants, saves the answers and counts what is in them. The counts become the metrics described in how to measure AI visibility: mention rate, citation rate and share of voice. What you pay for is the number of prompts, the number of assistants and how often each prompt is run.
This lesson explains how the tools work, how to judge one and what the vendors’ own pages listed on 4 October 2026. It does not rank vendors. Our AI visibility tools comparison lists them by category.
How the tools work
Every tool in this field follows the same four steps.
- Send prompts. The tool submits a prompt to an assistant, through the public app interface or through the model’s API.
- Store the answer. It saves the text, the linked sources and, where the assistant exposes them, the searches it ran.
- Extract. It identifies brand names in the text and domains in the sources.
- Aggregate. It turns those into rates over time, per assistant and per competitor.
Step 1 hides the main difference between products. An answer fetched through an API can differ from the answer a person sees in the app, because the app may add search, memory and location. Ask each vendor which method it uses for each assistant.
Two designs
Prompt panels. You write or approve a list of prompts, and the tool runs them daily or weekly. Peec AI, Otterly, Profound and Scrunch work this way. The numbers are specific to your market, and they are only as good as your prompt choice.
Prebuilt prompt databases. The vendor runs a very large set of prompts and lets you search the stored answers. Ahrefs describes its Brand Radar product as covering more than 449 million monthly prompts, modelled from searches in its keyword database. Semrush offers a similar index. You get competitor and market data on day one, but the prompts are the vendor’s, not your buyers’.
Several products now do both. BuzzStream’s State of Digital PR 2026 survey of more than 150 practitioners found that 52% use Ahrefs Brand Radar, which suggests PR teams favour the database design for campaign research.
What they measure
| Output | What it tells you | Check before trusting it |
|---|---|---|
| Mention rate or visibility score | How often answers name your brand | How many runs sit behind each figure |
| Citation or source report | Which domains and URLs answers link to | Whether mentions and citations are reported separately |
| Share of voice | Your mentions against competitors’ | Who chose the competitor set |
| Sentiment | Whether the description is positive | It is machine-scored. Read a sample yourself |
| Prompt or “AI search” volume | How often a prompt is asked | It is modelled. No platform publishes real volumes |
| Fan-out queries | The hidden searches behind an answer | Available for some assistants only |
| Crawler or agent analytics | Which AI bots visit your pages | Needs log or CDN access |
The source report is the most useful output for off-page work. It shows the third-party pages that assistants cite for your prompts, which is the list you work through in link building for AI visibility.
What they cost
We opened each vendor’s own pricing page on 4 October 2026. The vendors below are examples in alphabetical order, not a ranking. These are list prices on monthly billing, in the currency each page showed, and they will change.
| Vendor page | What it listed on 4 October 2026 |
|---|---|
| Ahrefs Brand Radar | In pounds sterling. Custom prompts “starts at £40/mo”. The AI Visibility Index is listed at £159 a month. Ahrefs subscribers get a small free prompt allowance |
| Otterly | In US dollars. Lite $29 a month for 15 prompts, Standard $189 for 100, Premium $489 for 400. ChatGPT, Google AI Overviews, Perplexity and Copilot included. Gemini, AI Mode and Claude are paid add-ons |
| Peec AI | Four plan names, Starter to Enterprise. The page did not show prices when we fetched it |
| Profound | A free seven-day trial with 50 prompts and a custom-priced Enterprise plan. No self-serve price was shown |
| SE Visible, from SE Ranking | In US dollars. Basic $99 a month for 200 prompts, Core $189 for 450, Plus $355 for 1,000. Covers ChatGPT, Perplexity, Gemini, AI Overviews and AI Mode |
| Semrush | In euros. The pricing page, viewed from the EU, listed an SEO plus AI Search Starter plan at €190 a month on monthly billing, with 50 prompts tracked daily |
Two things stand out. Price per prompt is not comparable across vendors, because a “prompt” may be run once a day on four assistants or once a week on one. And the headline price often excludes assistants you want. On Otterly’s page, adding Claude to the Standard plan was listed at $109 a month.
Third-party roundups put entry plans at roughly $29 to $130 a month and mid-range plans at $190 to $500. They disagree with each other and with vendor pages, so use them only as a rough guide.
The free layer
Before paying, set up what costs nothing.
- Bing Webmaster Tools has had an AI Performance report since February 2026. It shows Copilot citations by page and the grounding queries behind them.
- Google Search Console counts AI Overviews and AI Mode in its Web totals.
- GA4 has an AI Assistant channel. See AI referral traffic.
- Manual checks. Run 20 prompts by hand, five times each, and log the answers in a sheet.
The free reports cover Google’s and Microsoft’s own surfaces. They say nothing about ChatGPT, Perplexity or Claude answers. That is the gap a paid tool fills.
How to judge a tool
Ask every vendor the same questions.
- How many times is each prompt run, and how often? One run a week gives you one sample of an answer that changes every time.
- App or API? Per assistant, which does the tool query?
- Which assistants are included in the price, and what does each extra one cost?
- Are mentions and citations separate? They overlap less than you would expect.
- Can I see the raw answers? You need them to check accuracy and to audit the tool’s brand detection.
- Which countries and languages? An answer to an English prompt from a US location says little about Greece or Poland.
- How is “prompt volume” produced? If the answer is not “modelled”, be cautious.
- Can I export everything? You will want the cited URL list outside the tool.
Run a trial with 20 of your own prompts and compare the tool’s answers with what you see in the app. Many vendors list a trial of 7 to 14 days.
When a tool is not worth it
If you have fewer than a few dozen prompts that matter, a monthly manual check and the free reports are enough. If your brand is rarely discussed anywhere on the web, a tracker will show a flat line. Spend the money on coverage through digital PR first.
A tool also cannot tell you why a number moved. The research that connects actions to AI visibility is correlational and comes largely from the vendors of these tools, including several named here.
Where to go next
If you also want to follow coverage on news sites, forums and social platforms, read brand mention monitoring. For terms such as citation and share of voice, see the glossary.
Common questions
What does an AI visibility tool measure?
It measures how often a brand is named in AI answers to a set of prompts, how often a domain is linked as a source, which other brands appear and which third-party pages are cited. Some also score sentiment and log visits from AI crawlers.
How much do AI visibility tools cost?
Vendor pages we opened on 4 October 2026 listed self-serve plans in US dollars from $29 to $489 a month, with other vendors priced in pounds or euros and enterprise plans priced on request. Prices in this market change often, so check the vendor's page.
Are AI visibility tools accurate?
They record real answers, but each answer is one sample of an output that varies. SparkToro's January 2026 research found the same brand list came back less than 1% of the time. A tool is only as reliable as the number of runs behind each figure.
Can I track AI visibility without a paid tool?
Yes, in part. Bing Webmaster Tools shows Copilot citations, Search Console counts Google's AI surfaces and GA4 has an AI Assistant channel. You can also run prompts by hand and log the answers, which is slow but costs nothing.
What is the difference between a prompt panel and a prompt database?
With a panel, you choose the prompts and the tool runs them for you on a schedule. With a database, the vendor has already run millions of prompts and you search the stored answers for your brand and competitors.
