Ahrefs Brand Radar Review November 2026: From an AEO Consultant

Ahrefs Brand Radar is the better buy when the brand already lives in Ahrefs and the job is a small set of prompts you wrote, on engines you can name, paid for in checks. It is a weak buy when the board wants yesterday's Grok answer, or when someone treats the $199 index card as every platform plus a content program. On 8 November 2026 the pricing page showed Lite at $129 a month, Standard at $249, and Advanced at $449, with 150, 300, and 600 checks on AI prompts. Custom prompt packages started at $50 a month. The Brand Radar page showed an AI Visibility Index card at $199 a month and a line that reads $699 a month for all models.

Last updated - November 2026

I am Austin Heaton. Most of the reporting I set up is for fintech teams. The rest is B2B SaaS and other industries when the buyer question looks the same. A visibility login is worth opening when I can see the mention, the cited URL, and the sentence. A score I cannot open is not a plan.

This page is Brand Radar on 8 November 2026. It is not the suite comparison. That page is Ahrefs vs SEMrush for AI search reporting, and I am not pasting its prices forward as today's invoice. The Semrush side of a toolkit purchase is the Semrush AI Toolkit review. Prices and limits below are from pages I opened on 8 November 2026: Ahrefs Brand Radar, Ahrefs pricing, the help article How to set up custom prompts to track brand visibility in AI assistants, and the product post Grok in Brand Radar, higher API limits, and more (April 2026). The help article's byline said "Updated over 2 weeks ago" when I opened it. I am not turning that into a calendar day. A dedicated log with a different engine mix is the Otterly.ai review. The four-tool shortlist is Profound vs Peec AI vs Otterly vs Semrush.

Key Takeaways

• On 8 November 2026 the pricing page showed Lite at $129 a month, Standard at $249, and Advanced at $449, with 150, 300, and 600 checks on AI prompts, illustrated as 5, 10, and 20 prompts updated daily. The help article says every paid plan except Starter includes that allowance. Starter was $29 a month on the same pricing page.
• Custom prompt packages on the pricing page are Basic at $50 a month for 2,500 checks, Growth at $100 for 7,000, and Scale at $250 for 25,000. Overage is $0.020, $0.015, and $0.010 a check. The help article defines one check as one prompt, one model, and one location. The Brand Radar page says Claude consumes 8 checks per update, and that 5 prompts on 1 platform equal 150 checks a month.
• The visible index card says $199 a month. The same page prints "$699/mo for all models." The help article says the $699 all-platforms purchase includes 2,500 checks a month at $0.020 overage. A pricing configuration in the page source lists $199 a month for separate indexes named Google AI Overviews, ChatGPT, Gemini, Perplexity, and Grok, and $699 for all indexes. I am not merging those into one line on an order form.
• The Brand Radar FAQ names six index platforms: AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini, and Copilot, re-tested monthly on a 90-day window. Custom prompts add Claude and can run daily, weekly, or monthly. The help article still lists Grok and says Ahrefs cannot collect new Grok data for the time being. The April 2026 product post, bylined 25 May 2026, said Grok had been added.
• Brand Radar does not publish the page. The FAQ splits a mention (the brand name in the answer text) from a citation (a link to your page or domain). If the missing piece is a stored daily answer on an engine the quota will not cover, read the Semrush review before you add an index. A score is not a citation.

What Is Brand Radar, and What Is It Not?

Ahrefs sells Brand Radar as two ways to look at the same question. The page calls them Prompt Tracking and the AI Visibility Index. Prompt Tracking is the list you write. The card says it starts at $50 a month, and a line above the plan table says it is free for every paid Ahrefs Lite plan and above. The index is the database you search. The card says it starts at $199 a month, with no setup, and that it is best once the brand already appears in AI search. The page says the index covers 542 million prompts. The meta description on the same URL says "500M+ prompts from 6 AI tools." The help article, in a tip, says "350M+ prompts." I am not averaging three counts into one market size.

That split is the whole review. The index is useful when a known brand wants a map of topics it did not think to type. Custom prompts are useful when the question is specific, the brand is new, or the query would never show up in a keyword database. The FAQ says the index is built on Ahrefs' keyword database and may have limited coverage for brands with little or no search volume. Custom prompts are populated from the moment you start tracking. If a seed-stage product has no search demand yet, the index will not invent a history for it. The prompt list will, after the checks run.

Brand Radar is not the rest of Ahrefs, and it is not a writer. Site Explorer, Keywords Explorer, Site Audit, and Rank Tracker are the SEO subscription. The pricing FAQ says you can buy Brand Radar on its own, and that the purchase also includes an Ahrefs Free account. Content Kit, which the pricing page prices from $99 a month, is a separate add-on for AI Content Helper, an AI Content Grader, and a content inventory marked soon. I am not reviewing Content Kit here. A citation gap inside Brand Radar does not draft the page, and it does not publish it.

Two free-looking modules sit next to the product and should not be counted as the index. AI traffic in Web Analytics is marked free, with an install link. Bot visits are marked free while in beta. YouTube, TikTok, and Reddit monitoring is marked a bonus for Ahrefs customers, free while in beta. The FAQ says Brand Radar tracks social channels that influence AI answers, and it names Reddit threads ranking in Google, and YouTube. I did not install Web Analytics or Bot Analytics on 8 November. A crawler log is not the sentence in the answer.

The page also pitches Letaido for bulk prompts and dashboards. I did not price Letaido, and it is not part of this invoice.

How Do the Brand Radar Products Compare?

The table uses the Brand Radar page, the pricing page, and the help article, all read on 8 November 2026. Where a card, a help article, and a pricing object in the page source disagree, the cell keeps the disagreement. I am not picking the friendlier number.

If the included checks and the index are about to be two invoices, Book a Call and bring the prompt list and the engine list. The list shows whether Lite's 150 checks are a sample or the whole category.

What Does the AI Visibility Index Actually Count?

The FAQ is the definition I would put on a slide, and it is a definition of the index, not of a prompt you typed this morning. Ahrefs says Brand Radar runs 542 million or more search-backed prompts through AI platforms and stores every response. Visibility is then calculated from how often the brand is mentioned or cited. Four metrics are named. A mention is how often the brand name appears in the AI answer text. A citation is how often AI links to your pages or domains as a source. AI Share of Voice is the percentage of AI responses in the topic set that mention or cite your brand versus competitors. Estimated impressions are potential visibility weighted by the search volume behind each prompt. Share of voice, in that sentence, mixes a mention and a citation into one percentage. I would not add a mention total to a citation total and call it the same chart. If Monday's slide needs one number, I would label it with the metric name and the date I exported it.

The collection story is also on that FAQ, and it is two indexes, not one. Prompt-based indexes cover ChatGPT, Perplexity, Gemini, Copilot, AI Overviews, and AI Mode. Ahrefs says it takes queries from its keyword database, expands them through People Also Ask and what it calls semantic fanout, and runs the result through each platform. Search-query indexes cover AI Overviews and AI Mode only. Those track the keywords that trigger an AI answer inside Google, and whether the brand appears in it. They exist only for those two surfaces because, Ahrefs says, those are the only ones that show up in Google's own results. A prompt report and a keyword-trigger report are not the same row. I would not brief an executive that "we rank in AI Mode" when the chart is the prompt index and the question they asked was the Google trigger.

Freshness is monthly for the question sets, on a 90-day reporting window. That is the FAQ's sentence. Search Demand refreshes every few days. Web Visibility refreshes every few minutes. The index is therefore not a daily log of a prompt the sales team heard on Tuesday. It is a monthly re-test of Ahrefs' set, with faster refresh on the search-demand and web-visibility layers around it. Custom prompts are the daily object. Mixing a monthly index point with a daily prompt point on one line chart will look like a trend. It is two products.

The FAQ also says you do not set the index up. You can search a brand, a product, or a category immediately, including historical responses back to 2025, and there is no domain cap. That is the feature I would want for a brand that already has a name people search. It is the feature I would not promise to a company that renamed itself last quarter. History back to 2025 is Ahrefs' history of its prompt set. It is not a reconstruction of answers Ahrefs did not store.

The same FAQ says every index prompt maps back to search demand: the branded keywords behind the mentions, and topic clusters that show which pages get cited. If the buyer's question is not in the keyword database, it is not in this map. The help article's "350M+" and the meta description's "500M+" do not change that. I did not log in on 8 November. The custom-prompt help article says responses sit next to the rest of the dataset, with a filter for Ahrefs prompts, tracked prompts, or both. Confirm that control before a score goes into a deck.

How Do Custom Prompts and Checks Work?

Custom prompts are the product I would actually operate. The help article says you choose the assistants, the location, and the refresh. Tracked prompts can be monthly, weekly, or daily, and you can check only some of the assistants. The included allowance is on every paid plan except Starter: Lite 150 checks a month, Standard 300, Advanced 600. The pricing page prints the same three counts and adds the prompt illustration: 5, 10, and 20, with daily updates. Enterprise on that grid is "from 2,500" checks and "from 83" prompts. The Brand Radar page explains the illustration in one sentence: each prompt is checked daily on every selected platform, and 5 prompts on 1 platform equal 150 checks a month. Five times thirty is 150. The illustration assumes a 30-day month and one platform. It is not five prompts across every engine.

The help article's formula is stricter than the slogan. One check is one prompt execution, times one model, times one location. You can spend the checks across platforms, locations, and frequencies. A lower frequency leaves room for more prompts. The Brand Radar page adds an exception I would put on the order form in pen: Claude consumes 8 checks per update. I did not see that 8-check rule in the help article's formula. If both sentences are true, five Claude prompts, checked daily, are 5 times 8 times 30, which is 1,200 checks in a 30-day month. That is twice the Advanced allowance and eight times Lite. I did not run that configuration. I am not going to tell a Lite customer that Claude is included in any practical sense. It is available, and it is expensive in checks.

The same arithmetic changes the platform list. Five prompts on ChatGPT and Perplexity, daily, are 300 checks. That is the entire Standard allowance. Five prompts on the FAQ's six index platforms, daily, are 900 checks before Claude and before a second country. Advanced's 600 checks would not cover it. The marketing table's "5 prompts" on Lite is one platform, not a tour of AI search. A team that wants ChatGPT, Perplexity, and AI Overviews on the same ten buyer questions is already in a package, not in the allowance that comes with Lite.

Packages are on both the pricing page and the help article, and they match. Basic is $50 a month for 2,500 checks, overage $0.020. Growth is $100 for 7,000, overage $0.015. Scale is $250 for 25,000, overage $0.010. The pricing page's examples are 80 prompts daily on one platform for Basic, and 100 prompts daily on two platforms for Growth. Eighty times thirty is 2,400, which sits just under 2,500. One hundred prompts, two platforms, daily, is 6,000 checks, under the 7,000. The examples are illustrations of one way to spend the package, not a cap on prompt length. The help article says you can add one prompt per line, as long as you want, and the estimated checks sit above the button. Read that estimate before you save. Overage bills after the included checks are spent, if pay-as-you-go is on. If it is off, the help article says no further custom prompts are checked until the next billing month.

The page source on both the Brand Radar URL and the pricing URL also lists annual prices for those packages: $500, $1,000, and $2,500. That is ten times the monthly card, which is the same shape as "two months free," and it is close to the pricing page's "save up to 17%" line. I did not toggle an annual checkout for the packages. I am citing the configuration the page ships, next to the monthly cards I could read.

Setup is a saved report, then Track custom prompts. New data can take up to 24 hours. You can generate ideas from a location, one language at a time, and your own instructions, then edit them before saving. Deleting a prompt is not reversible. Tags group prompts. The help article says pulling custom-prompt data through the API is free and does not use API units. I did not call the API.

One sentence in the help article should not be skimmed. It says you can use the checks across ChatGPT, Gemini, Perplexity, and Copilot. The assistant list at the top of the same article also includes AI Overviews, AI Mode, and Grok. The marketing FAQ says custom prompts support the six index platforms plus Claude. I am not going to tell a client that every name on every list is selectable in the same dropdown. On day one I would open the selector and write down what it actually offers.

What Is the Status of Grok, as of 8 November 2026?

I opened the pages the question depends on, because a paused collector and a live SKU are not the same product.

The help article lists Grok as an assistant you can check custom prompts for, and then it says this: due to changes in Grok's policy, Ahrefs cannot collect any new data for the time being. The article had been updated more than two weeks before 8 November 2026. It does not say the historical rows were deleted. It says new data is not being collected. I am treating that as the operational note.

The product post titled as an April 2026 update is bylined 25 May 2026. It says Grok is available in Brand Radar for both Ahrefs prompts and custom prompts. It says Grok is included in the All Indexes add-on, or available separately at $199 a month. It says custom queries do not need that add-on. They need monthly checks. That was the announcement. It is not a sentence I can read on the Brand Radar FAQ today.

The Brand Radar FAQ's platform list, on the page I opened on 8 November, is AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini, and Copilot for the index. Custom prompts add Claude. Grok is absent from that list. The meta description's "6 AI tools" matches the six, not a seventh. Claude is the extra name, and it is a custom-prompt name, not an index name in that FAQ.

The pricing configuration inside the Brand Radar page and the pricing page still contains an object called indexGrok at $199 a month, or $1,990 a year. The same object lists index prices at $199 a month for Google AI Overviews, ChatGPT, Gemini, and Perplexity. I did not find index keys in that object for AI Mode, Copilot, or Claude. A SKU in the page source is not proof that new Grok responses are being stored this week. The help article is the sentence that talks about collection. The FAQ is the sentence that talks about what the product covers in the marketing copy. The configuration is the sentence that still knows how to charge $199 for a Grok index.

I am not going to guess that collection has resumed because the SKU is still there. I am also not going to guess that the SKU was removed, because I can see it. If the reason to buy Brand Radar is Grok, do not buy it on the strength of the May post. Ask Ahrefs whether new Grok responses are being stored, and ask to see one with a date after the help article's pause note. Until that row exists, Grok is a paused collector plus a price.

What the Invoice Says

The pricing page's three main cards, as I read them on 8 November, are Lite $129 a month, Standard $249, and Advanced $449. Annual billing "saves up to 17%," and those cards do not print a second annual dollar. Beside the check counts, Lite includes 750 tracked keywords and one user, with extra users at $40 a month. Standard includes 2,000 keywords, with extra users at $60. Advanced includes 5,000 keywords, with extra users at $80. I did not find a separate Brand Radar seat fee. A second person is an Ahrefs user on that card.

Starter is $29 a month and is listed apart from those three. The help article's exception is the sentence that matters: the custom-prompt allowance is on every paid plan except Starter. I would not buy Starter and expect 150 checks. Ahrefs Free is $0 and is not a prompt log. Enterprise is "talk to sales," with the check grid starting at 2,500 and the prompt grid starting at 83. There is no Enterprise dollar on the page.

The Brand Radar block on the pricing page says you can research any brand across 542 million or more organic prompts and track your own, and it says "get started from $199 a month." Under that, the custom packages match the help article, including the overage rates and the line that one prompt on one platform in one location uses one check. The standalone FAQ says Brand Radar can be the only tool you buy, and that you also receive an Ahrefs Free account. That is the path if the team does not want Site Explorer. The $199 starting price is then the index, and the $50 package is the prompt log. They are still two products. Buying the standalone index does not, on any sentence I read, include Lite's SEO history or 750 keywords.

The $699 line needs a careful read. On the Brand Radar page it appears as "$699/mo for all models," in the custom-prompts price block, before the index card's $199. The help article says the full all-platforms Brand Radar AI platform purchase is $699 a month and comes with 2,500 checks a month at $0.020 overage. The lines that follow the $199 in the page source are 83 prompts a day, plus 2,500 checks, the same $0.020 overage, "all platforms (Claude available)," and a custom refresh cadence. Those bullets would make the $199 card and the $699 purchase the same product, which they are not, if the help article is the contract. The page-source configuration keeps them apart: individual indexes at $199, all indexes at $699. I would put the engine names and the check allowance on the order form, and I would not sign a $199 order on the assumption that it includes every index and 2,500 checks. The help article attaches the checks to the $699 purchase. The card order on the marketing page is ambiguous. Ambiguous is not a discount.

Ahrefs says it does not run discounts, that there is no setup fee, and that refunds are not issued in general. I did not find a free-trial sentence for the index. The help article calls the included checks a built-in trial, which means an allowance inside a plan you already pay for, not a free week. I did not see the three-month pay-as-you-go carryover applied in writing to Brand Radar checks. I would not assume unused checks roll forward.

Which Limits Actually Change the Decision?

Five limits matter more than the feature tour.

The first is whose prompts they are. The index is Ahrefs' set, derived from the keyword database, People Also Ask, and fanout. It updates on a monthly question-set cycle. Custom prompts are yours, after you type them, and they can be daily. A weekly narrative about "the market" can come from the index. The fee question a buyer asked on Tuesday cannot, unless you tracked that sentence yourself. If the brand has no search volume, the FAQ says the index may not cover it. Start with prompts, and budget the checks.

The second is the check math. Lite's 150 checks are five daily prompts on one platform in a 30-day month, at one location. A second platform doubles it. A second country doubles it again. Claude multiplies an update by eight, on the Brand Radar page's rule. The packages exist because the included allowance is a sample. Basic's 2,500 checks are about 80 daily prompts on one platform, which is a real category for one engine, not a global program. Scale's 25,000 checks are the first package where a multi-engine, multi-country list starts to fit. Do that multiplication before the demo, not after the annual invoice.

The third is the engine that is actually collecting. The index copy names six platforms and does not name Grok. Custom prompts add Claude and, in the help article, a paused Grok. The May post said Grok was live. The page source still prices a Grok index. I would not put Grok on a slide this week. I would also not promise AI Mode and Copilot as separate $199 indexes, because I did not find those keys in the pricing object. They are named in the FAQ. A name in a FAQ is coverage. A key in the price object is a SKU. They are not the same list.

The fourth is the metric. A mention is the name in the answer. A citation is the link. Share of voice, as written, counts responses that mention or cite. Estimated impressions weight that by search volume, which brings the keyword database back into an AI chart. None of those numbers is the sentence. If the export or the response view is missing on the account you bought, do not brief the executives on the score.

The fifth is the page. Brand Radar does not write it and does not ship it. Content Kit is another subscription, from $99 a month on the pricing page, and I did not open its scoring rules on 8 November. Web Analytics can show which pages already receive AI traffic, if you install it. That is a clue. It is not a draft. The work still has to be published by someone who can stand behind a fee table or a product claim. On a live account, Austin Heaton still reads the answer before it goes into a client deck.

Honest Verdict: Who Should Buy, and Who Should Skip

Pick from the prompt, the engine, and the subscription you already pay for.

• You already pay for Ahrefs Lite or above, and the job is a handful of prompts on one engine. Use the included checks before you buy anything else. Lite is five daily prompts on one platform if you spend the month that way. Confirm the response view in the first 24 hours.
• The prompt list is the category, not a sample. Price a package from the check formula, not from the word "unlimited." Basic is the one-engine daily list. Growth is the two-platform illustration on the pricing page. Scale is the first tier with a lower overage. Ask which engines the selector actually runs before you multiply.
• The brand already has search demand, and you want topics you did not think to type. That is the index. Do not sign the $199 card until Ahrefs says, in writing, whether that dollar is one index or all platforms, and whether 2,500 checks are included. The help article puts the checks on the $699 purchase.
• You need Grok answers collected this week. Skip Brand Radar for that job. The help article says new data is not being collected. The May post and the $199 SKU do not overrule that note. Read the Semrush AI Toolkit review for a different engine list, and do not assume Semrush covers Grok either. That review says where Perplexity sits and where it does not.
• You are on Starter, or you wanted Brand Radar because it sounded included. Starter is outside the allowance. The index starts at $199 on the pricing page. The prompt package starts at $50. Neither one is inside the $29 plan.
• You need the article, not the chart. Brand Radar will not publish it. Content Kit is a separate add-on. A dedicated visibility log, if the engine list is the reason to switch, is the Otterly review and the four-tool shortlist.
• Fintech, payments, payroll, lending, or a B2B SaaS product with a fee table. Prefer the report that shows the sentence next to the source. An impressions number weighted by keyword volume does not check a license or a price. The index is a map. The prompt is the claim.

A tool subscription and an engagement are different bills. This page is only the tool. If the choice is a person rather than a login, the comparisons live on the compare page.

What I Could Not Confirm

• That the $199 index card includes every platform and 2,500 checks. The help article attaches those checks to the $699 all-platforms purchase. The marketing page prints both prices, and the bullets in the source sit under the $199.
• A single prompt-database size. The Brand Radar FAQ says 542 million or more. The meta description says 500 million or more. The help article says 350 million or more.
• That new Grok responses are being stored. The help article says they are not, for the time being. The May 2026 post said Grok was added. The page source still prices indexGrok at $199 a month.
• That AI Mode, Copilot, or Claude are separate $199 index SKUs. I did not find those keys in the pricing object. Claude is described as a custom-prompt platform that consumes 8 checks. AI Mode and Copilot are named in the FAQ.
• An annual dollar for Lite, Standard, or Advanced. The cards showed $129, $249, and $449, and the page said annual billing saves up to 17%.
• A logged-in response, a SERP snapshot, or an API call on 8 November 2026.
• Whether unused Brand Radar checks roll into the next month. The three-month carryover sentence I read was about other pay-as-you-go credits.

Frequently Asked Questions

What is Ahrefs Brand Radar?

Brand Radar is Ahrefs' AI visibility product, and the marketing page splits it in two. The AI Visibility Index searches a pre-built prompt database with no setup. Custom prompts track questions you type, on a schedule you choose. It is not Site Explorer, and it is not a content editor. Content Kit is a separate add-on.

Is Brand Radar included in a normal Ahrefs plan?

A monthly allowance of custom-prompt checks is included on every paid plan except Starter, according to the help article. The pricing page prints that allowance on Lite, Standard, and Advanced. The index is a separate starting price on both the Brand Radar page and the pricing page. Starter does not include the allowance. You can also buy Brand Radar on its own, and Ahrefs says that purchase includes an Ahrefs Free account.

How does Ahrefs count a mention, a citation, and a check?

The Brand Radar FAQ defines a mention as the brand name in the answer text, and a citation as a link to your page or domain. Share of voice counts responses that mention or cite the brand. A check, in the help article, is one prompt, one model, and one location. The Brand Radar page adds that Claude consumes more than one check per update. I would not add a mention total to a citation total.

Which engines can I track?

The index FAQ names AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini, and Copilot. Custom prompts on that FAQ add Claude. The help article's assistant list includes Grok and says new Grok data is not being collected for the time being. I am not merging those lists into one dropdown.

How fresh is the data?

The index FAQ says question sets are re-tested monthly on a 90-day reporting window. Search Demand refreshes every few days, and Web Visibility every few minutes. Custom prompts can be daily, weekly, or monthly. Newly added prompts can take up to 24 hours to show a response.

Who should buy Brand Radar, and who should skip it?

Buy the included checks if you already pay for Lite or above and the job is a short prompt list on one platform. Buy a check package when that list is the category. Consider the index when the brand already has search demand, after the starting price and the all-platforms price are reconciled in writing. Skip it when the board question is new Grok data, or when the missing piece is the page itself.

Will Brand Radar get the brand cited?

No. The reports describe mentions, citations, share of voice, and gaps. They do not publish the page. A check is a measurement. A citation is a URL in an answer, and someone still has to ship the page the model can repeat.

Read next:

• Ahrefs vs SEMrush for AI Search Reporting
• Semrush AI Toolkit Review November 2026
• Profound vs Peec AI vs Otterly vs Semrush
• Otterly.ai Review November 2026
• Goodie AI vs Profound for AI Visibility
• Compare AEO consultants and agencies

If the check math, the index price, and the paused engine still point at three different answers, Book a Call and we will put the prompt list, the platform selector, and the $199 versus all-platforms line on one page before anyone adds a package.