Head to head
Profound vs
Ahrefs Brand Radar
read off both pages.
We make a tool that competes with both of these, so read this knowing that. Every figure below was read off Profound’s and Ahrefs’ own pages on 31 August 2026, and each one links back to the page it was read from. Nothing here came from a review site, a roundup, or memory.
There is nothing to buy from us, so this page is not trying to close you on anything. It ends with a section on where each of them beats us — including the one thing Ahrefs published first that we had wanted to call ours — and a list of what we wanted to compare and could not.
Reviewed 3 September 2026 Vendor figures read from tryprofound.com and ahrefs.com on 31 August 2026 · a few capability lines carry forward from 30 August, where the vendor page had not changed · prices in this category move, so open the links before you buy anything
The short answer
Which of these is
even your question?
These two are not the same kind of product, and choosing between them starts there. One is a platform built only for this, with a writing layer attached. The other is a feature inside a tool a lot of marketing teams already pay for.
Profound
A platform built only for AI answers, with a writing layer.
Their cheapest plan is $99 a month, billed for a year with two months free. On that plan it watches ChatGPT and nothing else, tracks 50 prompts, and includes 100 “Agents credits” a month plus email support. The Agents are the part that writes content for you.
The next plan up is $399 a month for three answer engines, 100 tracked prompts and 400 credits. Above that is a custom-priced tier listing up to nine answer engines, several companies in one account, and SSO/SAML with SOC2.
Ahrefs Brand Radar
A feature inside an SEO tool, plus a checker anyone can open.
Custom Prompts start at $50 a month — their page says “starts at”, and does not print a ceiling for that add-on. The full AI Visibility Index is $199 a month for 83 prompts a day and 2,500 checks a month, with extra checks at two cents each. Their checker page shows the same ladder as $199 a month for selected platforms and $699 a month for all of them.
There is also a checker on their site that costs nothing to use. It asks six assistants, states “No signup required. Results in seconds”, and shows a limited preview — the top five results per section.
Side by side
What the money
actually buys.
Four columns, because these are the four things that change the answer. Every figure links to the page it was read from.
| Tool | Cheapest paid plan | AI assistants watched for that | Separates “named you” from “linked to you” | Hands you the fix |
|---|---|---|---|---|
| Profoundtryprofound.com | $99 / monthBilled for a year, two months free. Next plan up is $399 a month; the tier above that is custom-priced. | 1 on the $99 plan — ChatGPT only.3 on the $399 plan. Up to 9 on the custom tier. The wider list is what the platform can do, not what the entry plan watches. | No. “Citations” is used throughout with no stated separation.Carried forward from 30 August. | Yes. Their Agents generate content, paid for in credits.What one credit buys is not verified — their page says a credit is consumed each time an Agent runs, and never says what a run produces. |
| Ahrefs Brand Radarahrefs.com | from $50 / monthFor Custom Prompts — their page says that price starts there. The full AI Visibility Index is $199 a month, with 83 prompts a day, 2,500 checks a month, and overage at $0.020 a check. | 6 on the checker — ChatGPT, Google Gemini, Perplexity, Microsoft Copilot, Google AI Overviews and Google AI Mode.The paid ladder on that page runs $199 a month for selected platforms, $699 a month for all of them. | Yes, and they define both in plain words on two live pages — one of which costs nothing to reach. | No. It tells you where you stand and stops there.Carried forward from 30 August. |
| Answermendthat’s us | Nothing to buy yet.There is no paid plan and no checkout. What is open costs nothing and takes no account. | 2 — ChatGPT and Gemini, the pair whose cost per call we have actually measured.Google’s AI Overviews is on no plan here: it has no public API, and we would rather call it unmeasurable than scrape it and dress that up as a measurement. | Yes, and it is the reason this business exists — but it is a rule we follow, not a thing only we do. See the section below. | In part, today: the generators hand you the block of text or the small file to paste in. |
Where they part company
Four differences that
change the decision.
Not features lists. The four places where the two products answer a buyer’s question differently.
One of them writes. The other reports and stops.
Profound sells a fix layer from its cheapest plan: Agents that generate content, metered in credits, with a credit consumed each time an Agent runs. Ahrefs Brand Radar is diagnosis. It will tell you where you stand across six assistants and it will not write anything for you.
That is the single largest difference on this page, and it is the one that decides who you should be talking to. It is also worth knowing what Profound does not say: their pricing page never states what one Agent run produces, so “100 credits a month” is 100 runs, not 100 finished pages.
tryprofound.com/pricing · ahrefs.com/brand-radar · read 31 August 2026
How many assistants does each one watch?
They run in opposite directions. Profound’s $99 plan watches ChatGPT and nothing else; three assistants arrive at $399, and up to nine only on the custom tier. Ahrefs’ checker asks six with no plan at all, and their paid ladder sells selected platforms at $199 a month against all platforms at $699.
Breadth is not the same as depth, and neither page tells you how deeply a given assistant is sampled. But if the count is what you are buying, the two are not close at the entry price.
tryprofound.com/pricing · ahrefs.com/ai-visibility-checker · read 31 August 2026
Only one of them tells you what a citation is.
Ahrefs states both, verbatim, on their Brand Radar page: “Mentions – how often your brand name appears in AI answer text” and “Citations – how often AI links to your pages or domains as a source”. Their checker’s FAQ goes further: “A mention means the AI platform named your brand in its response. A citation means it linked to your website as a source. Both matter, but they require different strategies. Brand Radar tracks both separately.”
Profound uses “citations” throughout with no stated separation. Being named from a model’s memory and being read and linked to are two different failures needing two different fixes, so a single blended number hides the one you would act on.
ahrefs.com/brand-radar, /ai-visibility-checker · read 31 August 2026
Where do the questions come from?
Ahrefs makes a claim about this that is worth taking seriously: “Unlike tools that use synthetic (made-up) questions, Ahrefs derives all prompts from real search behavior — questions people actually ask on Google.” Profound makes a different one, about where the answers are read from: “responses directly from the consumer experience, not API outputs” — a harder thing to collect than the developer connection, and closer to the screen a customer looks at.
Both are checkable claims rather than slogans, and we are repeating them here rather than arguing with them. Reading answers off the consumer surface is harder than what we do. And Ahrefs publishes the cost of their own method on the same page: because the index is built on their keyword database, it “may have limited coverage for brands with little or no search volume”. That is a vendor telling you before purchase which brands its product measures badly, and there is no version of this page where we get to skip it.
ahrefs.com · tryprofound.com · the Profound line carried forward from 30 August
Where we lose
Where each of them
beats us.
A comparison page from a vendor that never loses a row is worth nothing to you. Here is every row we lose to these two.
Both of them can be bought today.
Ours cannot. There is no paid plan, no checkout, no sign-up. If you have a decision to make this month, we are not on the shortlist, and no amount of the rest of this page changes that.
Ahrefs got to the honest metric first.
Being named in an answer and being linked to as a source are two different problems. We think that distinction is the most important thing in this category, and we had planned to present it as ours. Ahrefs already defines both, in plain words, on two live pages — one of which anyone can open without paying. So it is a rule we keep, not a thing only we do, and we are not going to pretend otherwise.
Ahrefs is already in the stack, and asks for no account.
A feature inside a tool a team already pays for beats a new monthly bill for a lot of people. And their checker states “No signup required” in writing, with no cap on how many brands you can look up. We had counted “see it before you pay” as ours. It is the category standard.
Profound is built for a buyer we are not built for.
Seats, a security review, a vendor questionnaire, SSO/SAML, SOC2, several companies in one account. That is their end of the market and it is not ours. If that is your purchasing process, we are the wrong answer and we would rather say so here than waste your afternoon.
Profound reads answers off a harder surface.
Their own words: responses captured from the consumer experience rather than from API outputs. That is more work than reading a developer connection, and it is closer to the screen your customer is actually looking at. We read the developer connection.
Six assistants against our two.
Ahrefs’ checker asks six. Profound reaches up to nine on its custom tier. We ask two, ChatGPT and Gemini, because those are the two whose cost per call we have measured against a live key. We are not going to win a breadth race we did not enter.
The blanks
What we could not
check, and left blank.
A blank is more use to you than a guess. Here is each thing we wanted to compare, and the reason it is not in the table.
- 01What one Profound Agents credit actually buys. Their pricing page says a credit is consumed each time an Agent runs, and does not say what a run produces — a drafted page, an optimisation pass, one API call. So “100 credits a month” is a number we can repeat and not a quantity we can compare. It closes from their documentation or a trial, and until it does, the cell says not verified rather than guessing at a value.
- 02Whether Profound’s report puts a wall up after you submit it. Their AEO report form has exactly one input, placeholdered “Enter your website URL”, with no email, name or company field on the form itself. What happens after the submit button is unverified, because submitting a form on another company’s site is not something we are going to do on your behalf. Anyone can settle it in ten seconds by typing a domain in.
- 03How many times either one asks a question before reporting a number. Ask an AI assistant the same thing five times and you can get five different answers, so a figure from a single ask is close to meaningless. Neither company says what it does here. Nor do we, so here it is: our own check asks each question once rather than averaged, and one pass is not a trend.
- 04Anything at all about how the two compare on results. We have no customers, no case studies and no before-and-after to show you, so there is no outcome column and there is not going to be a fabricated one. Every claim on this page is about what these two companies publish, on pages you can open yourself.
The other side of this pairing is the price gap rather than the product gap: Profound against Otterly puts the same platform next to the cheapest tool in the category. The full eight-tool comparison has the rest.
Or skip the comparison
and just look.
Our own check asks one brand, ten questions, two assistants, no account and no email.