- Profound lists Starter at $99/month and Growth at $399/month, both billed yearly, plus a custom Enterprise tier with no public price.
- Starter tracks ChatGPT on 50 unique prompts, one seat, 100 Agent credits a month, and no exports. Growth adds Perplexity and Google AI Overviews, 100 unique prompts, three seats, 400 Agent credits, and CSV/JSON export.
- The complaint we hear most about Profound is that tracking prompts is expensive. The public plans also cap unique prompts at 50 or 100, and a small set leaves most buyer questions unmeasured.
- A visibility report is only useful if you can turn findings into published pages. A lot of teams do not have a plan for that work, and it is slower than buying the tool.
- API access, SSO, Prompt Volumes search data, ChatGPT Shopping, and tracking more than three answer engines are only included in Enterprise.
If you're shopping an answer engine optimization (AEO) platform, Profound is one of the names that comes up first. It sells visibility tracking across AI answers, plus agents for content work, and it now publishes self-serve prices instead of sending every buyer to a demo.
The public prices are listed as $99 or $399 a month, billed yearly. Here's what those numbers include as of August 27, 2026, which limits they come with, and what else you need in place for the subscription to change what you publish.
Profound's published prices
Profound describes itself as a full-stack platform for understanding, analyzing, building, and measuring brand visibility in AI search. Select “For brands” on the pricing page and you get three cards:
Starter is $99/month and Growth is $399/month, both billed yearly, each with a "2 months free" note. Enterprise is custom: tailored packages, no public dollar amount. There is no month-to-month toggle, and the cards don't spell out the annual charge, so the amount you actually pay shows up at checkout or on a sales quote.
On October 21, 2025, Profound said these plans start at $99/month and replace Lite. Existing Lite customers could stay or move. The live FAQ still asks "Should I use Lite or Enterprise?" so some Lite language remains.
The comparison table underneath is long. These are the rows that actually change:
Starter is ChatGPT tracking on 50 unique prompts, one seat, 100 Agent credits, and no exports. Growth is the first published plan that covers more than ChatGPT: three engines, 100 unique prompts, three seats, 400 credits, and CSV/JSON export. Both stay on one language, one region, daily runs, and email support.
API access, SSO/SAML, Prompt Volumes searches, ChatGPT Shopping, Projects, more than three seats, and engines beyond ChatGPT, Perplexity, and Google AI Overviews are Enterprise. That column lists capability for Claude, Gemini, Copilot, Grok, DeepSeek, and Google AI Mode as well.
The page also has a For agencies toggle. Profound's Agents FAQ names a self-serve Agency Growth plan with 400 credits per month per client workspace. The public brand cards we pulled don't show a dollar amount for that agency base.
How Agents credits work
Profound Agents are priced in credits. Each run consumes credits relative to complexity: fewer steps, fewer credits; multi-step agents, more.
Starter includes 100 credits a month. Growth includes 400. Agency Growth includes 400 per client workspace. Anything above those thresholds requires an Enterprise plan.
Profound Sheets is included on Starter and Growth, so what runs out first is the monthly credit allotment. Before a run, the platform shows an estimated cost. After a run, actual credits hit the balance.
If you go over, the account can be set to keep running with overage billing or to pause. There isn't a single public overage rate. Prompt limits are separate: Profound says it will notify you as you approach the cap and discuss raising it.
Setup, per the FAQ, is instant. The ongoing work is choosing which prompts are worth using one of the 50 or 100 unique-prompt allowances.
Prompt tracking is expensive, and a small set is less useful
The complaint we hear most often about Profound is that tracking prompts is expensive. That matches how the product is sold. You pay on an ongoing basis for each unique prompt you track: Starter includes 50, Growth includes 100, and anything larger is custom. Teams that already use the tool run into those caps when they try to add topics. Profound's own FAQ says more prompts mean more accurate insights, and that it will discuss raising the cap when you approach the limit.
You still want a large prompt set if the panel is going to represent how buyers actually ask. A visibility score is produced by sending targeted queries to models and counting how the brand appears. We have described that method as practical, and also as a prompt panel rather than first-party query volume. It reports on the questions you chose to run, not on every question people type. You need an adequate breadth of queries before you know what to cover, and paid tools like Profound are one way teams try to measure prompt volume.
Fifty or 100 unique prompts can be a starting sample. It is a small sample if you sell to several personas, sit in a category with many related questions, or need to score mentions, recommendations, and citations separately on the same answers. That pass only works on the responses you actually collected. If the prompt list is short, or skewed toward a few topics, the dashboard can look complete while most of the buyer conversation is unmeasured.
Cutting the number of prompts lowers the bill. It also means the report covers fewer of the questions buyers actually ask, which is usually why the team bought the tool.
You still need a way to act on the data
The other gap is what happens after the report. Profound can show where you appeared, which pages were cited, and which competitors showed up. That is a research input. It is not a plan for what to write, who reviews it, or how it gets to the site.
In calls, the constraint we hear is rarely a shortage of ideas from the dashboard. It is whether the team can turn those ideas into published pages. Getting copy approved, through subject-matter-expert review, and live on the site is slower than generating another visibility report. If that process is not in place, a larger prompt set produces more findings you cannot act on. That is extra cost for data that sits unused.
Buying the tool does not install the process for acting on the data. You still have to brief the page, get it reviewed, publish it, update it later, and measure whether it did the job you assigned. The loop is: monitor LLM visibility, find the gaps, publish to close them, and watch the answers again. You start with analytics, then decide which topics to prioritize. The pages you publish are what change what models say about your brand.
A lot of teams buy measurement first and assume the "what do we do with this" step will appear later. In practice it often doesn't, because it is a different kind of work: assigning owners, mapping questions to URLs, and getting copy through review. Budget for that labor next to the subscription. SEO and AEO often share about 6–10% of the marketing budget, and most of that money goes to strategy, production, and keeping pages current, not only a visibility platform.
Paid visibility tools sit next to free first-party sources such as Google Search Console and Bing's AI Performance report. Profound can measure a prompt set. It doesn't replace the content those models are reading, and it doesn't replace the process of turning a finding into a published page.
Conclusion
Profound publishes two self-serve prices: $99/month and $399/month, billed yearly. Enterprise is priced in a sales quote. Starter watches ChatGPT answers on a 50-prompt cap. Growth tracks three answer engines, with exports and three seats. Wider engine coverage, API, SSO, shopping, and prompt-volume searches require a custom plan.
The prompt, engine, and seat limits on each plan decide whether the subscription covers the monitoring you want to run. Tracking more prompts costs more, and a short list measures less of the market. The subscription is also only one part of the spend: you still need a way to turn the report into published pages, which many teams do not have mapped out when they buy the tool.
Frequently asked questions
Is Profound billed monthly or yearly?
The public Starter and Growth cards show a monthly price billed yearly, with a "2 months free" note. There is no month-to-month toggle on the page we reviewed on August 27, 2026. Enterprise is a tailored quote. Confirm the annual charge in checkout or with sales, because the cards don't spell out the invoice math.
What is the difference between Starter and Growth?
Starter is ChatGPT tracking, 50 unique prompts, one seat, 100 Agent credits, and no exports. Growth adds Perplexity and Google AI Overviews, 100 unique prompts, three seats, 400 Agent credits, and CSV/JSON export. Both stay on one language and one region, email support, and no API or SSO.
Does Starter include Perplexity or Google AI Overviews?
No. The comparison table puts Perplexity and Google AI Overviews on Growth. Starter is listed as "ChatGPT tracking only." Broader engine lists (Claude, Gemini, Copilot, Grok, DeepSeek, Google AI Mode) appear on Enterprise.
Why is tracking prompts expensive, and how many do you need?
Each unique prompt is billed on an ongoing basis, and the public plans stop at 50 (Starter) or 100 (Growth). Expansion is a conversation with Profound. A small set is cheaper, but the panel only measures the questions you run. For the report to represent how buyers ask, you generally need more prompts than a trial list, especially across personas and related topics.
What happens if we exceed prompt or credit limits?
For prompts, Profound says it will notify you as you approach the cap and help you look at expansion options. For Agent credits, the account can be set to continue with overage billing or to pause when the allotment is reached. Credit usage is visible before and after each run.
What do we need besides the dashboard?
A way to act on the findings: which pages to write or update, who reviews them, and how they get published. Without that, a larger prompt set mostly produces more unused findings. That work lives in a content system (research, production, measurement), not in the visibility tool itself.
When does Enterprise make sense?
When you need more than three seats, more than three engines, an API, SSO/SAML, Prompt Volumes searches, ChatGPT Shopping, Projects, or dedicated Slack support. Profound's FAQ still says teams of more than three people are better suited to a custom enterprise plan. Enterprise is also what you need if you want a prompt volume well above the public caps.
How do Profound Agents credits work?
Credits are consumed each time an agent runs, scaled to complexity. Starter includes 100 per month, Growth 400, and self-serve Agency Growth 400 per client workspace. Higher thresholds require Enterprise. Profound Sheets is included on Starter and Growth. The public brand cards do not show an agency dollar amount; ask Profound for that price list.
Related reading
- How to budget for AEO and SEO — Where an AEO tool subscription sits inside the 6–10% SEO/AEO budget, and why most of the spend is still production and the content system.
- How do you build a content system tied to pipeline, not just the blog calendar? — How to turn research findings into briefs, reviews, published pages, and measurement.
- What's the difference between being mentioned by an LLM, and recommended? — How to read mention, recommendation, and citation as separate signals instead of one blended visibility number.
