Twelve tools now track brand visibility in AI answers, and they differ on three axes that decide fit: how many engines they query, how often they requery, and whether they explain absence or only report it.
Entry prices run from $29 to $399 a month. Only two of the twelve publish a permanent free plan, and only three publish how many times they re-run a prompt, which is the number that decides whether their figures mean anything.
We sell one of these twelve, so read this with the scepticism that deserves. What we can do is make it checkable: every figure below was read from the vendor’s own pricing page on 9 August 2026, each one is linked, and where a vendor does not publish something this page says so rather than guessing.
What actually separates these tools?
Feature lists in this category are close to identical and close to useless. Three things genuinely differ, and each one changes what the numbers mean.
How many answer surfaces get queried, and whether the ones you care about are in the tier you can afford.
Entry tiers here range from one engine to nine.How many times the same prompt runs per engine per day. Answers vary between identical runs, so this decides whether a figure is a rate or a coin flip.
Only three of twelve publish it.Whether the product tells you that you were missing, or why. The first is a dashboard, the second is a work queue.
Most of this category does the first.What kind of tool are you actually buying?
The twelve are not twelve versions of one product. They fall into three shapes, and the shape matters more than any feature, because it determines what the tool is optimised to be good at.
Measurement is the whole product. Rivarise, Otterly, Mentions.so, Peec AI, Profound, Trakkr, Knowatoa, Rankscale. Deepest coverage per dollar and nothing else in the box.
Eight of twelveMeasurement plus content actions and attribution. Goodie, AthenaHQ, Scrunch. More expensive at entry, and they expect to be a programme rather than a dashboard.
Three of twelveAI visibility sold alongside an SEO platform. Ahrefs Brand Radar, and Semrush sells one too. Both are separately priced add-ons rather than free extras, and Brand Radar in particular is shaped around an index of observed prompts rather than a prompt set you choose.
One in this listWhat do the twelve cost?
Entry tier, because that is the one most teams actually buy. Every figure links to the page it came from.
| Tool | Entry price | Engines at entry | Prompts at entry | Free plan |
|---|---|---|---|---|
| Otterly.AI | $29/mo | 4 core | 15 | No permanent free plan |
| Mentions.so | $49/mo | 3 LLMs | 25 | No permanent free plan |
| Knowatoa | $59/mo | ChatGPT, AI Overviews, AI Mode | Not published | Free trial, no permanent free plan |
| Rivarise | $69/mo | 4 | 50 per day | Yes, permanent, no card |
| Rankscale | $99/mo | Your choice, credit-metered | 1,200 credits | 7-day free trial on Pro |
| Profound | $99/mo billed yearly | ChatGPT only | 50 | No public free plan |
| Trakkr | $100/mo | All 8, included | 50 per brand | 14-day free trial |
| Scrunch AI | $250/mo Core | Not stated on the pricing page | Not stated on the pricing page | No public free plan |
| AthenaHQ | $295/mo Starter | 9 models | 3,600 credits | Yes, Essential tier with $25 of credit and 300 credits |
| Ahrefs Brand Radar | $398/mo Select Platforms | AI Overviews and AI Mode, ChatGPT, Copilot, Gemini, Perplexity, Grok | An index of observed prompts rather than your own set | Firehose brand-mention tool free; Brand Radar itself is not free |
| Goodie AI | $399/mo Explorer | 3 core | 100 prompts, 3,000 responses | 7-day free trial |
| Peec AI | Not published | Six named models | Credit-metered | No permanent free plan listed |
Sorted by entry price, with vendors that publish no price listed last. “Not published” means the vendor’s own page does not state it, which is a finding rather than an omission on our part.
Rivarise and AthenaHQ. Everything else is a trial or a demo, which means you evaluate against a countdown rather than against your own data.
Rivarise, AthenaHQ and Goodie state runs or response budgets. Everywhere else, “daily tracking” could mean one run or fifty, and the difference decides whether the number is real.
Why do the prices not compare cleanly?
Because three different meters are in use, and a headline price tells you almost nothing about what you will actually spend.
Most of this list. Easy to compare across vendors and easy to size, as long as you also know how many times each prompt is re-run. Without that second number, “50 prompts” is not a quantity of anything.
Rankscale and AthenaHQ. More honest about the underlying cost and considerably harder to compare, because your monthly spend depends on choices you make after you sign.
A third meter hides inside the first. Several vendors sell the engines separately from the prompts, so the tier price covers three or four models and the rest arrive as add-ons. Otterly bills engines beyond its core four separately, and Peec meters models as a multiplier rather than bundling them, though it has stopped publishing what any of it costs. Neither is hidden, and both mean the number in the pricing table is a floor rather than a price.
A tool that meters models as a multiplier is not quoting you a price, it is quoting you a starting point. Peec documents one prompt against one model for a month at 30 credits, so six models costs exactly twice what three does, and Otterly bills its extra engines on top of every tier. It is the same purchase with the arithmetic left to you.
Which one fits which situation?
In entry-price order. Each entry states where the tool is weaker, including ours, because a comparison where one product wins every row is a sales page wearing a comparison’s clothes.
The cheapest way to start tracking anything at all.
- Where it is weaker
- Engine add-ons and extra prompt packs mean the sticker price is rarely the real one.
- Requery frequency
- Daily tracking stated, runs per day not published
- Top self-serve tier
- $489/mo for 400 prompts, engines beyond the core four billed as add-ons
Agencies wanting many sites on one bill, with unlimited seats at every tier.
- Where it is weaker
- Prompt allowances are low at the bottom two tiers relative to the price.
- Requery frequency
- Not published
- Top self-serve tier
- $399/mo Agency, 300 prompts, unlimited sites
Reporting-stack teams: the API, MCP and Looker Studio integrations arrive at $199 rather than at enterprise.
- Where it is weaker
- Question allowances are not on the pricing page, so you cannot size it before talking to them.
- Requery frequency
- Daily updates stated, runs per day not published
- Top self-serve tier
- $199/mo Growth, all 7 services, plus API, MCP and Looker Studio
Teams that want all nine engines on one plan without add-ons, and the full answer text behind every number.
- Where it is weaker
- No longer the cheapest way in, and it does not do content generation or optimisation actions.
- Requery frequency
- 5 runs per prompt per engine per day, published
- Top self-serve tier
- $349/mo Business for all 9 engines and 80 prompts a day; $1,299 Agency for 5 client brands
European teams wanting country-level segmentation and unlimited users on every tier.
- Where it is weaker
- No tier carries a public price as of 30 August 2026, so you cannot size or compare it without a sales conversation.
- Requery frequency
- Daily tracking stated, and its documentation puts 1 prompt against 1 model for 1 day at 1 credit
- Top self-serve tier
- Advanced, up to 5 projects. No price published on any public page
Teams that would rather buy a budget than a plan, and audit pages as well as track prompts.
- Where it is weaker
- Credit maths makes cross-vendor comparison genuinely hard, which may or may not be deliberate.
- Requery frequency
- Credit-metered rather than fixed; you decide the run count
- Top self-serve tier
- $780/mo Enterprise, 12,000 credits
Enterprises that need SSO, SOC 2 and multi-company tracking, and the best-known name in the category.
- Where it is weaker
- The entry tier is one engine. Meaningful coverage starts at $399 with three.
- Requery frequency
- Not published
- Top self-serve tier
- Enterprise, custom, up to 9 answer engines and SSO/SAML with SOC 2
Anyone tired of per-model add-on pricing. Every model is in every tier.
- Where it is weaker
- One brand at the entry tier, so multi-brand work moves you up quickly.
- Requery frequency
- Daily per brand, runs per day not published
- Top self-serve tier
- Agency tiers above, with client portals and brand comparison
Brands that want an agent-facing view of their own site alongside visibility tracking.
- Where it is weaker
- The public pricing page states a price and very little else, so evaluation needs a call.
- Requery frequency
- Not published
- Top self-serve tier
- Enterprise, custom
Teams who want a free tier to evaluate with and full model coverage the moment they pay.
- Where it is weaker
- The jump from free to $295 is the steepest on this list.
- Requery frequency
- Credit-metered, 1 credit = 1 AI response
- Top self-serve tier
- Enterprise, custom, adds SAML and OIDC SSO, audit log and multi-region
Category research. The prompt index answers what people are actually asking, which no prompt set of yours can.
- Where it is weaker
- A separate add-on on top of your Ahrefs subscription, and the top tier is the one with custom prompt checks.
- Requery frequency
- Reports on an index of observed prompts, so a per-prompt run count does not apply in the same way
- Top self-serve tier
- $699/mo All Platforms, including 2,500 custom prompt checks a month
Teams buying optimisation actions and revenue attribution, not only measurement.
- Where it is weaker
- Two of the three tiers are demo-only, so you cannot self-serve past the entry plan.
- Requery frequency
- 3,000 AI responses a month at entry, which is a published run budget
- Top self-serve tier
- Enterprise, quote only, up to 11 models and 500+ prompts
What does none of them do?
An answer that names you and sends no click is invisible in analytics. Every vendor claiming otherwise is inferring.
There is nothing to submit and nothing to buy. Tools measure the odds, they do not set them.
The heaviest lifting is third-party evidence on domains you do not own. No dashboard does that work for you.
How should you choose?
One last piece of advice that costs nothing. Whatever you buy, write your prompt set down before you see any vendor’s dashboard, and do not change it for a quarter. The tool matters less than holding the prompt set and the run count constant, and a fixed set on a cheap plan produces a more defensible baseline than a drifting set on an expensive one.