Where to find AI prompt volume
People ask whether Profound or Evertune shows real prompt volumes. Here is what prompt-volume data is, how each vendor sources it, and where the caveats lie.
Search for AI visibility tools and you will find a question that comes up again and again: which one shows real prompt volumes? Buyers ask whether Profound or Evertune is better for it, whether AthenaHQ's numbers can be trusted, and whether anything exists that works like keyword volume in classic SEO. This guide explains what prompt-volume data is, how the main vendors source it, where the numbers are measured and where they are modelled, and how to use them alongside a tracker.
We read each vendor's own website on 7 November 2026. Vendors change their products often, so confirm details before you buy. We have left out prices, which are on each vendor's pricing page.
What prompt-volume data is
In classic SEO, keyword volume tells you roughly how many times a month people search a phrase. Prompt volume is the AI equivalent: an estimate of how often people ask an AI assistant a given question, or a question with the same intent.
It is harder to get. Google publishes search data through its own tools. AI assistants do not publish what users ask them. Anyone selling prompt volume therefore has to get the data some other way, and the method decides how far you can trust the number.
It helps to separate this from tracking. A tracker runs a list of prompts that you chose and records what the AI says about your brand. Prompt volume tells you which prompts are worth putting on that list. One is about answers, the other about demand. Our piece on how to measure share of voice in AI search covers the tracking side.
The three ways vendors source it
Almost every approach falls into one of three groups:
- Panel data. A vendor gets prompts from real users of AI assistants who have agreed to share them, then scales the sample up to the wider population.
- Modelled estimates. A vendor builds a model, often from search-engine data and other signals, and predicts how often a prompt is asked.
- Query expansion. A vendor takes known search demand and turns it into natural-language prompts that people are likely to ask.
Each has a different failure mode, which the sections below set out.
Profound Prompt Volumes
Profound's Prompt Volumes page says it licenses prompts from double opt-in panels of real AI assistant users. It says the data is anonymised and aggregated, and the result is statistically modelled to correct demographic and geographic bias and to estimate broader-market volumes. It describes hundreds of millions of prompts a month.
The page says data is updated weekly and that history begins in January 2025. Its FAQ lists ten countries: the US, Canada, Italy, Brazil, Germany, Australia, Spain, South Korea, France and the UK. It classifies prompts by intent (informational, commercial, conversational or generative), and lets you filter by platform, age range and income bracket.
Note the wording: panel data, then statistically modelled. The base is measured, but the headline figure is still a scaled estimate. Profound's pricing page lists five Prompt Volumes searches on its free Trial, and custom volume on Enterprise. See our Profound alternatives guide for the wider picture.
Evertune EverPanel
Evertune's homepage says its proprietary EverPanel holds over 150 million user prompts, and that its Prompt Agent uses the panel to find the questions consumers ask, and group related prompts into topics. Its pricing page lists Prompt Volumes and Prompt Lab on Pro and Enterprise, not on the self-serve Starter plan.
We could not find a public methodology page that explains how the panel is recruited or weighted, so questions such as the countries covered and the refresh cadence are worth putting to Evertune directly. For a head-to-head on this feature, read Profound vs Evertune, or our list of Evertune alternatives.
AthenaHQ prompt-volume estimation
AthenaHQ's pricing page lists unlimited prompt volume estimation and analysis, with numerical search estimates, on its Starter and Enterprise plans. Its Query Volume Estimation Model, which it calls QVEM, is described as a proprietary machine-learning system. The vendor says it draws on public data partner data and proprietary data, and that its ingestion layer collects from over 50 sources. It says the model outputs estimates with confidence intervals, but does not specify how they are calculated.
This is a modelled estimate rather than a panel. That is not necessarily worse, but it means the number is only as good as the model, and the public description does not let you check it. See Evertune vs AthenaHQ and Pineprompt vs AthenaHQ.
Otterly.AI prompt research
Otterly's prompt research tool is not a panel. Its help centre says Otterly estimates "intent volume" with an algorithm based on Google search volume, because AI platforms do not disclose their usage data. It states plainly that this is an estimate and not a directly measured count of AI searches, and it shows the result as a five-level indicator from very low to very high.
The tool itself generates prompt suggestions grouped by intent, funnel stage and topic. Otterly's pricing page lists it on some higher-tier plans. It is a useful way to build a prompt list, and an honest one about its limits. See Pineprompt vs Otterly and our Otterly alternatives.
Ahrefs Brand Radar
Ahrefs describes search-demand estimates for AI prompts in Brand Radar. As the vendor explains it, the data starts from Ahrefs' keyword database: real queries are expanded into natural-language prompts using People Also Ask and semantic fan-out, run through AI platforms, and the responses are stored. Search demand refreshes every few days. That is query expansion, tied to Ahrefs' keyword search data rather than to what people typed into an assistant. See Pineprompt vs Ahrefs Brand Radar and our explainer on prompt fan-out.
What we could not verify
Semrush's AI Visibility Toolkit documentation that we read did not say whether it offers prompt-volume data or how it would be sourced, so we make no claim either way. Scrunch's homepage describes AI Search Trends with approximate prompt volumes, but does not say where the data comes from. We did not find prompt-volume features on the pricing pages of Peec AI or Promptwatch, although Peec lists prompt suggestions on every plan. If you rely on one of them for demand data, ask the vendor.
The caveats: measured versus modelled
Whatever the source, keep these in mind:
- No one sees all AI usage. Panels cover the users who opted in. Models cover whatever data they were trained on. Both are samples.
- Panels are modelled too. Scaling a panel to a population needs weighting, and weighting is a modelling choice.
- Prompts are long and varied. People phrase the same question in many ways, which makes exact-match counting hard. Volume by intent or topic is usually more stable than volume by exact wording.
- Treat figures as relative. Use them to rank topics against each other, not as precise counts.
- Ask for the method. A vendor that explains its sources, as Profound and Otterly do, is easier to evaluate than one that does not.
How to use prompt volume with a tracker
- Use volume data to choose your prompt list: topics with real demand, in the right country and language.
- Put those prompts into a tracker and run them daily on the AI platforms your customers use.
- Compare visibility against demand. A topic with high volume and low share of voice is your priority.
- Re-check the volume source every quarter, since AI usage shifts quickly.
If tools give you conflicting numbers, our article on why AI visibility tools disagree explains why.
Where Pineprompt fits
Pineprompt does not have prompt-volume data. It is a focused tracker. It runs daily on five live AI platforms: ChatGPT, Gemini, Claude, Google AI Overviews and Google AI Mode. Grok is listed but temporarily unavailable. Each monitor covers one country and one language. The MCP server is on every plan, and the API and Looker Studio connector are on Professional and Business. There is no free trial, and no content agents, ads or audits.
So if you need demand data, source it from one of the vendors above or from your own search data, then bring your chosen prompts to Pineprompt to see how AI platforms answer them. See the pricing page and the comparison with Evertune.
Common questions
Is Profound or Evertune better for seeing real prompt volumes?
Both build on real-user panels. Profound's page describes double opt-in panels with weekly updates across ten listed countries. Evertune describes EverPanel with over 150 million user prompts. Profound says it models its panel to estimate wider volumes; we could not find a public methodology page for Evertune. Neither is a complete count, so ask each for a sample of your category before choosing.
Does ChatGPT publish prompt volume?
We did not find prompt volume published by AI assistants themselves. As Otterly's help centre notes, AI platforms do not disclose their usage data, which is why vendors use panels or estimates.
Is AI prompt volume accurate?
It is an estimate. Panel-based figures are scaled from a sample, and modelled figures depend on the model. Use them to compare topics, not as exact counts.
Does Pineprompt show prompt volume?
No. Pineprompt tracks how AI platforms answer the prompts you choose, and does not provide prompt-volume data.
Track your brand across every AI platform.
Pineprompt monitors five live AI platforms daily, in the country and language your buyers use. See our methodology for what we capture and how.