How to track AEO without a subscription

There are four ways to track AEO without a monthly subscription: script it yourself against the provider APIs, run the free open-source or free desktop tools that already exist, buy a one-time lifetime deal for a cloud tool, or buy a one-time desktop app like AI Visibility Tracker for $29. All four share the same economics — you pay AI providers directly for the queries at cost instead of paying a platform to mark that usage up. Which one is right depends almost entirely on how much of your own time you want to spend maintaining it.

Last updated 2026-07-28

What are the actual non-subscription options?

Four, and it is worth naming them honestly, including the ones that cost nothing and might mean you do not need to buy anything at all.

  • Build it yourself. A script that loops your prompts across provider APIs with web search enabled, stores results, and computes a rate. Entirely doable in a weekend and free apart from API usage.
  • Free open-source tools. geo-aeo-tracker is an MIT-licensed, local-first, bring-your-own-key project that tracks several engines and keeps data in your browser's storage. Early-stage but real, and genuinely free.
  • Free desktop utilities. Ahoy Lighthouse is a free macOS app that runs local bring-your-own-key AI visibility audits across a range of platforms, with keys kept on your machine.
  • One-time purchases. Lifetime deals for hosted AI-visibility tools appear periodically on marketplaces like AppSumo — one-time pricing, but the software is still cloud SaaS and the deal terms are the vendor's. AI Visibility Tracker is one-time and local: $29 once, no subscription.

Should you just build it yourself?

If you are comfortable with an API and you want to understand the measurement deeply, yes — and you will learn more doing it than reading anything on this site. The core loop is genuinely simple: for each prompt, for each engine, call the API with web search enabled, read the structured citations, check the answer text and citations for your brand terms, store the result, divide.

The parts that take longer than expected are the parts that make the numbers trustworthy rather than merely present. Grounding has to be enabled correctly per provider and each returns citations in a different shape. Brand matching needs word boundaries, diacritic handling and domain normalisation, or "Notion" matches "notional" and you spend a week debugging a metric. Failed calls must be excluded from the denominator rather than counted as misses. Costs need capping before a loop bug spends real money. Scheduling needs to survive a sleeping laptop. And you need a trend to be worth anything, which means running it every day for months.

That is the honest pitch for buying something: not that the loop is hard, but that the maintenance is boring and never ends. If that is a good trade for $29, buy it; if you would enjoy the build, build it.

Are the free tools good enough?

Sometimes, genuinely — and anyone in this category who tells you otherwise is protecting a price. If you want to run an occasional audit rather than a daily trend, a free tool that queries a set of engines and reports where you appear will answer the question at zero cost.

The differences to check before deciding are the ones that determine whether the numbers are stable enough for a trend: does it run on a schedule unattended, or only when you click it? Does it store history in something durable, or in browser storage that a cache clear removes? Are citations read from provider metadata or parsed out of prose? Is the measurement model fixed, or does it change as the project updates? Are failed calls excluded from the denominator? Those are the same questions to ask us — the answers are on why web-grounded measurement is non-negotiable.

What do you actually pay in API costs?

A few cents per day for a typical core prompt set, and that is the whole ongoing cost of every non-subscription option on this page. A grounded call costs more than a plain completion because it carries search fees and reads retrieved content, so the arithmetic is roughly: prompts × engines × per-call cost, once a day.

For ten to fifteen core prompts across five engines that lands in the low tens of cents daily, or single-digit dollars a month — an order of magnitude below the cheapest hosted tier, because you are paying the provider rather than a platform reselling the same calls. AI Visibility Tracker shows a projected cost before each run and enforces a per-run budget cap, defaulting to $5, that skips a run projected to exceed it. Runs that would duplicate the same day's work are deduplicated rather than re-billed.

What each non-subscription route costs and demands
OptionSoftware costOngoing costYour time
Build it yourselfFreeAPI usage at costA weekend to build, then ongoing maintenance as provider APIs change
Open-source trackerFreeAPI usage at costSetup, plus whatever the project does not yet do
Free desktop utilityFreeAPI usage at costMinimal for audits; check whether it schedules and stores history
Lifetime deal on a cloud toolOne-time, varies by dealIncluded in the dealMinimal — but the data and the terms are the vendor’s
AI Visibility Tracker$29 once, no subscriptionAPI usage at cost, capped per runInstall, add a key, define prompts once

What does a subscription buy that none of these do?

Managed infrastructure and organisational fit — and it is worth stating plainly because it is the honest case against everything else on this page. A hosted platform runs your prompts at a fixed time regardless of whether any machine of yours is awake, absorbs every provider API change, gives colleagues a URL rather than a file, and produces the compliance artefacts a procurement team asks for.

It also buys datasets that are structurally unavailable to a local tool: AI crawler analytics, agent traffic attribution, multi-region answer simulation and vendor-curated industry benchmarks. If you need those, no amount of not paying a subscription will get them. The full picture of who sells what is in AI visibility tools compared.

So when is a $29 one-time app the right answer?

When you want a daily trend rather than an occasional audit, you do not want to maintain a script, and you do not need anyone else to have a login. That is a narrow and real set of people: founders, solo marketers, consultants and small teams tracking one brand or a handful.

If you need seats, compliance or crawler data, buy a platform. If you only want to check occasionally, use a free tool and keep the $29. If your prompt set and competitor list are commercially sensitive, local storage is the argument that matters more than the price. AI Visibility Tracker includes a 7-day free trial, which is the cheapest way to find out which of those you are.

Frequently asked questions

Is there a free AI visibility tracker?

Yes. geo-aeo-tracker is an MIT-licensed open-source local-first project, and Ahoy Lighthouse is a free macOS utility that runs local bring-your-own-key audits. Both require you to supply your own AI provider API keys, and neither is a managed service — check whether they schedule unattended runs and store durable history if you need a trend rather than a one-off audit.

Can I track AI visibility with a spreadsheet?

For a one-off audit, yes — ask each engine your prompts manually and record whether you appeared. It stops working as a measurement almost immediately, because answers vary run to run, so a hand-collected sample cannot separate a real change from ordinary variance. Anything you intend to read as a trend needs to be automated and run daily.

Do I need a subscription to track AEO?

No. Every non-subscription route — a script you write, a free open-source tracker, a free desktop utility, or a one-time app like AI Visibility Tracker at $29 once — runs on your own provider API key and pays for queries at cost. Subscriptions buy managed infrastructure, seats and compliance rather than a fundamentally better measurement.

What does AI Visibility Tracker cost in total?

$29 once for the app, no subscription, plus your own AI API usage paid directly to the provider — typically a few cents a day for a core prompt set across five engines. Every install includes a 7-day free trial, and the app caps per-run spend at a budget you set, defaulting to $5.

Related guides