Ahrefs Brand Radar is Ahrefs' AI visibility product: it measures how often a brand is mentioned and cited in AI answers, using two separate data sources - a pre-collected AI Visibility Index of about 496 million search-derived prompts (from $199 a month) and Custom Prompts that run your own questions (from $50 a month for 2,500 checks). It is the strongest option on the market for researching any brand's AI visibility without setup, and a weaker fit if you need fresh, per-prompt chatbot data, Claude in the index, or help fixing what it finds.
This review is built from Ahrefs' own pricing page, help center, methodology post and changelog, all checked on October 24, 2026, plus a close read of 14 of the strongest pages ranking for Brand Radar queries, including Ahrefs' public feature-request board. You get what is really inside the prompt database (most of it is Google, not chatbots), verified prices with cost-at-volume math, the index-versus-custom-prompt trade-off explained with our own variance data, sourced limitations, a scorecard and alternatives.
Key takeaways
- Two products in one: the AI Visibility Index ($199/month per platform, $699 for all platforms) is pre-collected research data; Custom Prompts ($50, $100 or $250 packages) track the questions you choose. Prices checked October 24, 2026.
- The database is mostly Google: of 496.8 million monthly prompts on Ahrefs' product page, 77% are AI Overviews and AI Mode. Each of ChatGPT, Gemini, Perplexity and Copilot has about 28 million.
- Chatbot index data refreshes monthly: ChatGPT, Gemini, Perplexity and Copilot question sets are re-tested once a month on a 90-day reporting window. Custom Prompts can run daily.
- Claude is custom-prompt only and costs 8 checks per run. Grok collection is paused.
- You can buy it standalone: Ahrefs' pricing FAQ says Brand Radar does not require a paid Ahrefs SEO plan, so reviews quoting $828 to $1,148 a month are adding a base plan you may not need.
- Best for: competitive and category research, and Ahrefs users who want AI data next to SEO data. Less suited to monitoring a short list of buying questions closely or to closing the gaps it reveals.
What is Ahrefs Brand Radar?
Ahrefs Brand Radar is a tool for researching and tracking how brands appear in AI-generated answers. In Ahrefs' words it "tracks how any brand shows up in AI search", across Google AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini and Copilot, with Claude available for custom prompts (Ahrefs help center, September 2026).
It is easier to evaluate once you split it into its parts, because they are priced and refreshed differently:
- AI Visibility Index - Ahrefs' own prompt database, already run through each AI platform. You type in any brand, product or category and see mentions, citations, AI Share of Voice and Estimated Impressions with history back to 2025. No setup and no cap on domains.
- Custom Prompts - the questions you add yourself, run daily, weekly or monthly per platform and location. History starts on the day you add them.
- Search Demand and Web Visibility - branded search volume and web mentions from Ahrefs' keyword and crawl data.
- Beta channels - YouTube and TikTok mentions (titles, descriptions, transcripts) and Reddit results that rank in Google, free for paid Ahrefs customers while in beta and due to become paid add-ons afterwards (Ahrefs, December 2025).
The product has moved quickly. Ahrefs' changelog shows Claude arriving for custom prompts on July 13, 2026, AI-adjusted search volume in July, the chatbot prompt set doubling in August, and entity matching for ambiguous brand names on August 21. Reviews written before mid-2026 describe a noticeably thinner tool.
Where does Ahrefs Brand Radar get its AI data?
The index is built from search data, not from real chatbot conversations. Ahrefs takes queries from its keyword database, expands them into questions using Google's People Also Ask and a semantic fan-out system, then enters those questions into each AI platform's free web interface with its default model (Brand Radar methodology, updated February 2026).
The details matter for how you read the numbers:
- No personalization. Prompts run with no stored user data or prior context, and Ahrefs says "no normalization, personalization, pre-prompting, or filtering" is applied. Real users with chat history may see different answers.
- Location follows the keyword. Each prompt is asked in the same location as the keyword it was sourced from, and the locale mix mirrors Ahrefs' keyword database. Coverage is strongest in English.
- Mentions are string matches; citations are linked URLs. Since August 2026 you can match a brand term as an entity rather than plain text, which cuts false positives for brands named after common words.
- Hallucinated links are kept. The methodology states: "We do not filter out hallucinated or malformed links."
- Claude is different. It is tracked through the API (Claude Sonnet 4.6 with web search, per the help center), not a web interface, and only for custom prompts.
- Estimated Impressions is a model. Mentions are weighted by search volume behind each prompt; Ahrefs calls this "a modeling choice, not a measured relationship".
What is actually inside the 496 million prompts?
Most of the headline number is Google. We added up the per-platform counts on Ahrefs' Brand Radar page on October 24, 2026:
| Platform | Monthly prompts | Share of total |
|---|---|---|
| Google AI Overviews | 279,880,612 | 56.3% |
| Google AI Mode | 102,742,648 | 20.7% |
| Gemini | 28,710,507 | 5.8% |
| Copilot | 28,689,960 | 5.8% |
| Perplexity | 28,632,830 | 5.8% |
| ChatGPT | 28,111,675 | 5.7% |
| Total | 496,768,232 | 100% |
So 77% of the database is Google's AI surfaces, and each chatbot gets about 28 million prompts a month. That is still far more chatbot data than any brand could collect itself, but it changes how to read "496 million": if ChatGPT is your main channel, the relevant sample is about 28 million, refreshed monthly.
Ahrefs' own figures for the size have also changed fast and do not always agree. The December 2025 announcement cited 239 million prompts a month; the February 2026 methodology listed about 14.5 million per chatbot; the help center (September 2026) says both "405+ million" and "+350 million"; the changelog says the chatbot set doubled "from around 14M, to 30M prompts" on August 5, 2026. Reviews quoting 218 million (Profound, March 2026) or similar figures are working from older snapshots. Use the live page and note the date.
How often is Brand Radar data updated?
It depends on the index. Google's AI surfaces refresh every few days; the chatbot indexes refresh once a month; custom prompts refresh as often as you schedule them.
| Data | Refresh cadence (per Ahrefs) | History starts |
|---|---|---|
| AI chatbots in the index (ChatGPT, Gemini, Perplexity, Copilot) | Once a month, on a 90-day reporting window | May 2025 |
| AI Overviews and AI Mode in the index | Every few days, in line with Keywords Explorer | AI Overviews: August 2024 |
| Custom Prompts (all platforms, plus Claude) | Daily, weekly or monthly - your choice per platform | The day you add them |
| Search Demand | Every few days | 2015 |
| Web Visibility | Every few minutes | 2013 |
| YouTube / TikTok / Reddit in Google (beta) | Not stated | December 2023 / December 2024 / 2015 |
Sources: Ahrefs help center and the Brand Radar page FAQ, checked October 24, 2026. A monthly refresh is fine for trend lines across thousands of prompts. It is slow if you just published a page and want to know whether ChatGPT now cites it.
Which AI platforms does Ahrefs Brand Radar track?
Six platforms in the index, plus Claude through custom prompts only. Coverage differs between the two products, which several reviews blur:
| Platform | AI Visibility Index | Custom Prompts | Notes |
|---|---|---|---|
| Google AI Overviews | Yes | Yes | Largest index; also a search-query-based index |
| Google AI Mode | Yes | Yes | Also a search-query-based index |
| ChatGPT | Yes | Yes | Fan-out queries shown when returned |
| Perplexity | Yes | Yes | Fan-out queries shown when returned |
| Gemini | Yes | Yes | |
| Microsoft Copilot | Yes | Yes | Few dedicated trackers cover it |
| Claude | No | Yes | 8 checks per run; added July 2026 |
| Grok | Paused | Paused | Added May 2026; help center says Ahrefs is "temporarily unable to gather new data" |
| Meta AI, DeepSeek | No | No | Not listed by Ahrefs |
Claude stays out of the index for cost reasons: the changelog says Claude "is only available to monitor for custom prompts, rather than the entire Ahrefs dataset". An "Add Claude Index" request on Ahrefs' Brand Radar feature board had 11 votes when we checked.
How much does Ahrefs Brand Radar cost?
Brand Radar starts at $50 a month for custom prompt tracking and $199 a month for one AI index; all AI indexes cost $699 a month and include 2,500 custom prompt checks. Paid Ahrefs plans also include a small custom-prompt allowance. These are list prices in US dollars from Ahrefs' pricing page and help center, checked October 24, 2026.
| What you buy | Price / month | What you get |
|---|---|---|
| AI Visibility Index, one platform | $199 | Pre-collected data for one AI index |
| AI Visibility Index, all platforms | $699 | All AI indexes plus 2,500 custom prompt checks |
| Custom Prompts: Basic | $50 | 2,500 checks; overage $0.020/check |
| Custom Prompts: Growth | $100 | 7,000 checks; overage $0.015/check |
| Custom Prompts: Scale | $250 | 25,000 checks; overage $0.010/check |
| Included in Ahrefs Lite / Standard / Advanced ($129 / $249 / $449) | - | 150 / 300 / 600 checks a month (5 / 10 / 20 prompts daily on one platform) |
A check is one prompt on one platform in one location for one run. Claude uses 8 checks per run. Prices vary by region: the same page displayed €47, €179 and €654 in a euro-priced session. Ahrefs did not show a free trial for Brand Radar; its free AI Visibility Checker gives a preview.
Can you buy Brand Radar without an Ahrefs subscription?
Yes. Ahrefs' pricing FAQ says: "Brand Radar is available as a standalone tool. When you purchase it, you'll also receive an Ahrefs Free account." TechRadar's October 2026 review reached the same conclusion. Several competitor reviews quote $828 to $1,148 a month because they add a $129 or $449 Ahrefs plan on top of the $699 index; that is the right total only if you also want Site Explorer, Keywords Explorer and the rest of the SEO suite.
What does tracking cost at realistic volumes?
Checks multiply by prompts, platforms, locations and runs, so the package you need is easy to underestimate. We worked out the cheapest standalone package for common setups (30-day month, one location, list prices above):
| Setup | Checks / month | Cheapest package | Monthly cost |
|---|---|---|---|
| 25 prompts daily on ChatGPT, Perplexity, Gemini | 2,250 | Basic | $50 |
| 20 prompts daily on 3 platforms in 2 countries | 3,600 | Basic + overage | $72 |
| 100 prompts weekly on 6 platforms plus Claude | about 6,060 | Growth | $100 |
| 50 prompts daily on all 6 index platforms | 9,000 | Growth + overage | $130 |
| 50 prompts daily on 6 platforms plus Claude | 21,000 | Scale | $250 |
| Research any brand across the full index | - | All platforms index | $699 |
The takeaway: custom prompt tracking in Brand Radar is cheap per check, Claude is the expensive platform, and the $699 index is the part you pay for research breadth rather than monitoring.
Index-style vs custom-prompt tracking: which do you need?
An index tells you how a brand shows up across a huge, fixed sample of market questions; custom prompt tracking tells you what happens to the specific questions your buyers ask. Brand Radar is unusual because it sells both, and most dedicated AI visibility trackers only sell the second.
| Criteria | Index-style (Brand Radar AI Visibility Index) | Custom-prompt tracking (Brand Radar Custom Prompts, AI Rank Lab, Semrush, Profound, Otterly) |
|---|---|---|
| Where prompts come from | Search queries expanded via People Also Ask and fan-out | You write or select them |
| Setup and history | None; history back to 2025 on day one | Setup needed; history starts when you add prompts |
| Chatbot refresh | Monthly | Daily, weekly or on demand |
| Reliability of one prompt's result | Low - one sample per month | Higher - repeated runs can be averaged |
| Reliability of aggregate share of voice | High - noise averages out over millions of prompts | Depends on how many prompts you track |
| New, niche or low-search brands | Thin coverage (Ahrefs says so) | Works from day one |
| Competitor research | Any brand, no setup, no domain cap | Only prompts you chose |
| Main cost driver | Number of platforms | Prompts x platforms x locations x frequency |
| Best for | Category research, competitor audits, agency pitches | Monitoring buying questions, measuring the effect of changes |
Why a single monthly snapshot of a prompt is noisy
AI answers change from run to run, so one sample per prompt per month tells you little about that prompt on its own. In AI Rank Lab's own data - 972 successful runs of 42 prompt-engine pairs, each run at least five times between February 13 and October 22, 2026 - 18 of the 42 pairs (42.9%) flipped between mentioning and not mentioning the tracked brand at least once. The flip rate was 23.1% on Gemini, 45.5% on Claude, 54.5% on ChatGPT and 66.7% on Perplexity. The median overlap in cited domains between consecutive runs of the same prompt was 0.41 (Jaccard), and only a median 3.4% of cited domains appeared in every run. The sample is small and covers a handful of tracked domains, so treat it as directional; our AI answer stability study describes the method.
This is not a flaw unique to Ahrefs. It is why index data works for aggregate questions ("does our category share of voice rise over a quarter?") and why individual buying questions need repeated runs. Ahrefs effectively says the same: its methodology calls the metrics "directional indicators, not exact traffic counts". If you calculate share of voice from either source, our guide to AI share of voice formulas and sample size shows how many prompts you need before a change means anything.
How accurate is Ahrefs Brand Radar?
Accurate for what it samples, and easy to misread. The most-quoted accuracy test, from Writesonic's review (updated January 2026), found Brand Radar reporting 3 ChatGPT mentions of Writesonic against 123 in Writesonic's own tracker. Competitor reviews repeat this as "a 97% undercount". But the two tools were counting different prompt sets: the index samples questions derived from search demand, while a brand's own tracker samples questions chosen because they are relevant to that brand. Neither number is "all mentions in ChatGPT". The test also predates the August 2026 doubling of the chatbot prompt set.
The fairer accuracy questions are narrower: are the stored answers real (yes, you can open each one), are mentions matched correctly (better since entity matching), and is the sample right for your market (check whether your key questions exist in the index before you buy). TechRadar's hands-on test found some reports "require manual verification", parts of stored answers showing raw markup, and a sidebar reporting no citations while listing pages as "found but not cited".
What does Brand Radar do well?
Its strength is research depth: you can move from a competitor overview to the individual AI answers and the pages behind them, for any brand, with history, on day one.
- Instant competitive research. No setup, no domain cap, historical answers back to 2025. No custom-prompt tracker can show you a competitor's AI visibility before you started tracking.
- Found versus cited. Since May 2026 the Cited pages report separates pages that appeared in a response from pages that were actually cited, and flags "found but not cited" brand mentions (Ahrefs, May 2026). That is a practical gap-finding view.
- Connection to search data. Each index prompt maps back to a keyword and topic, so you can see which high-volume topics trigger AI answers and which pages get cited.
- Copilot coverage. Microsoft Copilot is in the index and in custom prompts; many trackers skip it.
- Cheap custom checks. $50 for 2,500 checks works out at 2 cents per prompt run, falling to 1 cent on overage in the Scale package.
- Data access. API endpoints for reports, prompts and tags, a Looker Studio connector, Report Builder widgets, and Ahrefs MCP access on Lite and higher plans, which works with Claude, ChatGPT and Copilot (Ahrefs MCP guide, September 2026). Some Brand Radar MCP prompts need the matching index add-on.
- Adjacent channels. YouTube, TikTok and Reddit-in-Google visibility (beta) sit next to the AI data.
What are the limitations of Ahrefs Brand Radar?
The main limits are monthly chatbot data, search-shaped prompts, Claude outside the index, and no workflow for fixing what you find. Each is sourced below.
- Monthly chatbot refresh in the index. ChatGPT, Gemini, Perplexity and Copilot index data is re-tested monthly (help center). You need Custom Prompts for anything faster.
- Prompts reflect search behavior. Index prompts come from Google demand, so long, conversational, context-heavy prompts are under-represented. Ahrefs' methodology acknowledges that long-tail prompts "may be missing" and that coverage is strongest in English. Profound's review (a competitor) found no results for a niche query it tested.
- Thin for small or new brands. Ahrefs' own FAQ says the index "may have limited coverage for brands with little or no search volume".
- Claude only in custom prompts, at 8 checks per run; Grok paused. If Claude matters to your buyers, budget for it separately.
- Estimated Impressions are modeled, weighted by search volume, not measured AI usage. Ahrefs says it does not estimate actual AI usage by country.
- Sentiment is in transition. TechRadar saw a native Sentiment report in its review account in October 2026, while the "Sentiment analysis" request on Ahrefs' feature board was still marked Planned (16 votes). Older reviews saying "no sentiment" may be out of date; check your account.
- Feature requests that show real gaps: a date filter for Brand Radar (20 votes), tracking specific URLs (11) and exporting all AI responses for a time period, all open on the public board in October 2026.
- Hallucinated or malformed links are not filtered, per the methodology, so verify cited URLs before outreach.
- No fix workflow inside Brand Radar. It shows gaps; acting on them is up to you. Ahrefs' Content Kit add-on is listed "from $99/mo" and marked "Soon" on the pricing page, and its agent features now sit in a separate product, Letaido.
- Beta channels will be charged later. YouTube, TikTok and Reddit indexes are free only while in beta.
Ahrefs Brand Radar scorecard
Our overall rating is 3.6 out of 5: excellent for research, average for monitoring, weak for acting on results. This is an editorial assessment against Ahrefs' documentation and published tests, using the same criteria we apply to every AI visibility tool; we did not run a paid Brand Radar account for this review.
| Criterion | Score (out of 5) | Why |
|---|---|---|
| Competitive and category research | 5 | Any brand, no setup, history, stored answers and cited pages |
| Engine coverage | 4 | Six platforms plus Claude in custom prompts; Grok paused; no Meta AI or DeepSeek |
| Data freshness | 3 | Chatbot index monthly; custom prompts daily |
| Methodology transparency | 4 | Published methodology, but database-size figures conflict across Ahrefs pages |
| Price and value | 4 | Very cheap custom checks; $699 for the full index is a real budget line |
| Reporting and data access | 4 | API, MCP, Looker connector, Report Builder (add-on) |
| Helps you fix gaps | 1 | No content, audit-to-fix or publishing workflow in Brand Radar itself |
| Overall | 3.6 | Simple average of the seven criteria |
What are the best Ahrefs Brand Radar alternatives?
The right alternative depends on what Brand Radar leaves you missing: a fix workflow, an existing SEO suite, enterprise engine coverage, or the lowest entry price.
Disclosure: AI Rank Lab is our product. We list it first because this is our site, and we've kept the same criteria for every tool below.
| Tool | Entry price (monthly, list) | Engines | Helps fix gaps | Best for |
|---|---|---|---|---|
| AI Rank Lab | $69 Starter (10,000 credits, or 100 AI brand visibility checks if used only for that) | ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, Google AI Mode | Yes: SEO/AEO/GEO audit, Action Plan, content, publishing | Teams that want tracking and fixing in one place |
| Ahrefs Brand Radar | $50 custom prompts; $199 per index | Six in index plus Claude custom | No (Content Kit "soon") | Research at scale, Ahrefs users |
| Semrush AI Visibility Toolkit | $99 per domain, 25 prompts | ChatGPT, Google AI Mode and others by report | Recommendations and AI site audit | Existing Semrush users |
| Profound | Free 7-day trial; Enterprise custom | Up to 9 on Enterprise, including Claude, Copilot, DeepSeek | AI Marketer content drafts | Large and regulated brands |
| Otterly.AI | $29 Lite, 15 prompts | ChatGPT, AI Overviews, Perplexity, Copilot; AI Mode, Gemini, Claude as add-ons | GEO audit | Lowest entry price |
Prices checked October 24, 2026 on each vendor's pricing page.
AI Rank Lab runs the prompts you choose on ChatGPT, Claude, Gemini and Perplexity (brand visibility tracking), checks Google AI Overviews and AI Mode, records mentions, position, sentiment, competitors and cited URLs, and connects gaps to an SEO/AEO/GEO audit, an Action Plan and a content workflow that publishes to WordPress. It also connects GA4 and Search Console, and its MCP connector works from Claude. Plans are $69, $149 and $499 a month (pricing). Where Brand Radar is the better choice: per check, Brand Radar's custom prompt packages are much cheaper (2,500 checks for $50 versus 100 checks on our Starter credit pool), and only Brand Radar gives you a pre-collected index of any competitor's history plus Copilot coverage.
Semrush AI Visibility Toolkit costs $99 a month per domain with 25 prompts, according to the Semrush knowledge base. Our Semrush AI Visibility Toolkit review maps which engines each report covers. Where Semrush is the better choice: teams already living in Semrush who want AI data beside it.
Profound shows a free trial (50 recommended prompts run daily for 7 days on ChatGPT, Gemini and AI Overviews) and custom Enterprise packages with up to nine answer engines, SSO and SOC 2 on its pricing page. Where Profound is the better choice: enterprises that need the widest engine list and compliance.
Otterly.AI starts at $29 a month for 15 prompts on four core engines, with Gemini, AI Mode and Claude as paid add-ons (Otterly pricing). Where Otterly is the better choice: the smallest budget and a short prompt list.
For a wider field, see our ranked list of AI visibility tracker tools; for Ahrefs as a full SEO suite, see Semrush vs Ahrefs vs Moz.
Who should use Ahrefs Brand Radar?
Use Brand Radar if your main question is "how does our market look in AI answers?" Look elsewhere if your question is "did our changes move these 30 buying questions?" and you need help making the changes.
| Your situation | Recommendation | Why |
|---|---|---|
| Agency pitching or auditing many brands | Brand Radar index | Any brand, history, no setup |
| Already on Ahrefs, want a first look at AI data | Start with included custom checks | 150 to 600 checks are already in Lite to Advanced |
| Need Copilot plus Google AI surfaces at volume | Brand Radar | Strongest Google and Copilot coverage per dollar |
| New or niche brand with little search volume | Custom-prompt tracker | Index coverage is thin by Ahrefs' own account |
| Claude is a key channel | A tracker with Claude included, or Brand Radar Custom Prompts at 8 checks per run | Claude is not in the index |
| Need to close gaps, not just see them | A platform with audit and content workflow | Brand Radar has no fix workflow |
How to test Brand Radar before you commit
Run a two-week test that checks coverage first and spend second:
- List 20 to 30 buying questions your customers actually ask. Pull them from sales calls, support tickets and comparison searches.
- Use your included checks first. If you are on Lite or higher, add 5 to 20 of those questions as Custom Prompts on one or two platforms before buying anything.
- Search the index for your category (one platform, $199) and check whether your questions, or close equivalents, are in it. If they are not, the index will not answer your monitoring question.
- Open stored answers and confirm mentions and cited URLs by hand for a sample of ten. Switch on entity matching if your brand name is a common word.
- Run the same prompts repeatedly for at least a week and look at how often the result flips. AI Rank Lab automates this step by storing every run per prompt and engine, so repeated results can be compared over time.
- Price the setup you would really run with the scenario table above, including locations and Claude.
How we researched this review
We read Ahrefs' Brand Radar page, pricing page and FAQ, the help center article on Brand Radar, the methodology post, the changelog, the MCP guide and the public feature-request board, all checked on October 24, 2026. We then read 14 of the strongest pages ranking for Brand Radar queries, including TechRadar's hands-on review and reviews by competing vendors, and noted where they disagree with Ahrefs' current documentation. Most ranking reviews are published by companies selling alternatives; we flag that where we quote them, and the same applies to us. The variance figures come from AI Rank Lab's aggregated, anonymized prompt-run data (method and sample above). We did not have a paid Brand Radar account, so we did not run our own head-to-head accuracy test; the scorecard reflects documentation and published tests, not our own measurements of Brand Radar.
If you want your own prompts tracked across ChatGPT, Claude, Gemini, Perplexity, AI Overviews and AI Mode, with an audit and Action Plan to close the gaps, you can start with AI Rank Lab - signup includes 100 free credits, no card required.
Get a free AI ranking consultation
Want your brand to show up when ChatGPT, Gemini and Perplexity answer questions in your market? Tell us about your site and our team will send a personalized strategy.
