Profound vs Rankscale, the one pair where you can price a single AI engine before you buy

If the decision is going to be made on arithmetic, Rankscale is the only one of these two you can do the arithmetic for. Read on 31 October 2026, rankscale.ai/pricing publishes a credit weight against each engine it queries, and its FAQ states that a query typically costs 0.25 credits per engine per prompt. On the same day tryprofound.com/pricing publishes two plans for a brand, Trial at Free and Enterprise at Custom, and the engine list on the paid side reads Up to 9 Answer Engines tracked with no price attached to any of them. Both facts are honest positions. Only one of them lets a buyer work out next month's bill.

The published engine weights, which almost nobody in this category prints

These came from the credit calculator on Rankscale's own pricing page, read 31 October 2026.

Engine	Credits per prompt per run
Google AI Mode	0.25
ChatGPT	0.25
Perplexity	0.25
Gemini	0.25
DeepSeek	1
Anthropic Claude	2

The calculator also prints its own worked example, 50 prompts multiplied by 0.75 credits per run on a weekly schedule, giving 162.5 credits a month. Set that against the published allowances and the whole ladder becomes legible: Essentials carries 120 credits, Pro 1,200, Pro+ 3,000, Growth 5,500 and Enterprise 12,000, and the same table converts those into answers collected, 480, 4,800, 12,000, 22,000 and 48,000. Prices on that page are $20 a month for Essentials sold yearly only, then $99, $229, $385 and $780 monthly, with a yearly column at $17, $84, $195, $327 and $663.

What Profound publishes instead, and why it is not nothing

Profound answers a different question in print. Its Trial is free and gives 50 unique prompts run daily for 7 days across ChatGPT, Gemini and Google AI Overviews, with one language and one region. The comparison grid beside it is unusually frank about what the free tier withholds: History None, Exports None, API None, Support None, SSO No. Seats are published as unlimited on both plans, which is a real and uncommon commitment at the free end.

At Enterprise the grid lists SSO and SAML with SOC 2 compliance, all time history, exports in CSV and JSON, API access, custom languages and regions, ChatGPT Shopping, and support options up to a dedicated specialist with a 24 hour service level. Its FAQ answers the setup question in one word, Instant, and publishes the one allowance that carries a number on the agency side, 400 credits per month per client workspace on a self serve Agency Growth plan. So Profound publishes a scope of work and a governance commitment, and leaves the cost to a conversation.

Cadence is where the two plans diverge most

Rankscale's comparison table publishes scheduling frequency as hourly to monthly, and credit rollover of up to two times on its middle tiers and up to three times higher up. That is a tool built around the idea that you choose how often each question gets asked and carry unspent capacity forward.

Profound's free tier publishes exactly one cadence, daily for 7 days, and its Enterprise tier publishes daily. There is no published hourly option and no published rollover. For a brand in a fast moving category daily is the right answer and the question never comes up. For a buyer with 500 questions and a quarterly reporting cycle, the ability to run most of them monthly and a handful hourly is the difference between one plan and the next one up.

Which one to buy

• You need a number for a budget line this quarter. Rankscale. Its page prices five tiers, two billing terms and six engines individually, and you can model your own prompt set against it in ten minutes.
• Claude is on your must measure list. Read the weight before you commit. At 2 credits against 0.25 for ChatGPT, Claude costs eight times a standard engine check on Rankscale, so a Claude heavy prompt set changes which tier you need. Profound names Anthropic Claude in its Enterprise capability list and publishes no weight at all, so that cost is inside the quote.
• Procurement wants SSO, SOC 2 and an audit trail. Profound publishes all three on its Enterprise tier. Rankscale puts SSO on its unpriced Custom plan, alongside individual credits, custom API and service level agreements, so you reach the same place through a quote.
• You are an agency. Compare Rankscale's Growth tier, $385 a month with 50 brand dashboard slots, white labelled dashboard links and an agency programme that includes lifetime affiliate commissions and a partner directory entry, against Profound's 400 credits per client workspace. One is priced per container, the other per unit of work.

The weight question applied to us, including the part we lose

We sell in this category and we publish a per surface weight too, so the same comparison is fair to run on us. Our pricing page, read 31 October 2026, states that one credit is one buyer question asked on one AI app once, that most checks use one credit, and that a Claude check uses 3. Against a 0.25 engine that is a three times multiple where Rankscale publishes an eight times one, so on the specific question of what Claude costs relative to a standard check, we are the cheaper of the two and the figures are on both pages for anyone to check.

Two things go the other way and they are worth saying plainly. Rankscale publishes team workspaces with unlimited seats; our page publishes Logins included, 1, full access. Rankscale publishes a public changelog; we do not. It also publishes an Essentials tier at $17 a month on yearly billing, and the cheapest plan on our page that watches AI apps at all is GEO Growth at $299 a month, or Rs 11,799 in rupees. If budget is the binding constraint, Rankscale wins that comparison outright and we say so.

If Profound is the half of this pair you are seriously considering, the companion read is Profound alternatives on a budget or for a small team, which works through what a seven day window can and cannot settle.

Questions about pricing Profound and Rankscale against each other

Can I work out what Rankscale will cost me before signing up

Yes, and that is the main reason to shortlist it. Multiply your prompt count by the sum of the weights for the engines you want, multiply by how often you want them run, and compare against the credit allowance on each tier. The page does the same calculation in front of you with 50 prompts and 0.75 credits a run.

Why does Claude cost more than ChatGPT at both vendors

Because the vendors are paying for it differently and both say so with a figure. Rankscale prices Claude at 2 credits against 0.25 for its cheapest surfaces. Our own page gives the reason in one line, that a Claude check uses 3 credits because Claude costs more to run. A vendor that publishes no per engine weight has not removed that cost, only the visibility of it.

Does Profound's free trial include exports

No. Its own comparison grid lists History as None, Exports as None and API as None on the Trial plan, and all three as available at Enterprise. So readings taken during the seven days stay inside the product.

Which has the wider engine coverage

Rankscale's table says 17 or more AI engines, and its calculator lists named surfaces beyond the six weighted ones, including Copilot and several API model variants. Profound's Enterprise row says up to 9 Answer Engines tracked and names nine, including Exa Search, which nobody else in this set names at all.

Is the yearly only Essentials tier a catch

It is a condition, published clearly. Rankscale's comparison table marks Essentials as yearly only and prints $20 and $17 against it, so the cheapest entry to the product requires a twelve month commitment. Every other tier shows both a monthly and a yearly figure.

What should an agency compare first

Containers, not prompts. Rankscale publishes brand dashboard slots per tier, 2, 10, 25, 50 and 100, and says extra slots can be added in the app. Profound publishes credits per client workspace. Count your live clients, then read which number runs out first.

How we know this

We read the assistants rather than guessing what they do. In our own audit of 1,260 saved answers from ChatGPT and Gemini across the United States, the United Kingdom and India, 82.0% of every citation landed on a site the business itself owned, and review directories took 0.7%. Counted as one domain per reading, 16,516 of 20,143.

That is the whole asymmetry: the pages you own are the pages an assistant quotes back to a buyer. See the dated readings and the method on our proof page.

Find out whether assistants name your business, on your own domain and your own buying questions.

By Aditya Goel, Founder, AI Knows Us. Reviewed 1 November 2026.