AI visibility tools compared, with the criteria stated first

What AI visibility tools do, and the criteria used here

AI visibility tools send a set of buyer questions to assistants such as ChatGPT, Perplexity and Gemini, read the answers, and count which brands are named and which pages are cited. They differ in price, in how many assistants and questions they cover, in how often they check, and in whether they help you fix a loss.

We make Agentcites, which is one of the tools below. Every tool, ours included, is judged on the same criteria, and the criteria come first, before any tool appears. Every fact about another tool is read from that tool’s own page on 10 October 2026 and linked. Where a page does not say something, the table says "Not stated", which means we could not find it, not that the tool lacks it.

The seven tools side by side

The table puts the first five criteria in one place. The sixth, whether a tool writes the fix, has its own section further down, because the answers do not fit in a cell.

Each tool, read from its own pages

Each entry says what the tool’s own pages state, where it is ahead, and what we could not find. The order is the same as the table.

OtterlyAI

OtterlyAI’s pricing page lists Lite at $29 a month for 15 prompts, Standard at $189 for 100 and Premium at $489 for 400, with Enterprise from 1,000 prompts. Every plan includes ChatGPT, Google AI Overviews, Perplexity and Microsoft Copilot, and Claude, Google AI Mode and Gemini are paid add-ons. Tracking is daily. API access starts on Standard.

It is the cheapest paid plan in this list, and it covers Google AI Overviews and Copilot, which Agentcites does not ask. Its help centre describes its recommendations as a reasoning and a suggestion for you to act on, so the writing stays with you. Its home page mentions both a 7-day and a 14-day trial.

Peec AI

Peec AI’s instructions page for AI lists Starter at $95 a month for 50 prompts, Pro at $245 for 150 and Advanced at $495 for 350, each with three models of your choice and daily tracking. Its pricing page shows the same plans and prompt counts, lists Enterprise with "Up to 13 LLM models tracked", API access and single sign-on, and offers a free trial. The prices did not show in the page text we could read, so they are taken from the instructions page.

Peec AI tracks more prompts than Agentcites from its Pro plan up, tracks daily, and lets you choose which three models to watch. Its instructions page says it is "not a content-generation" tool "by design" and that it "deliberately does not write or publish content itself".

Semrush AI Visibility Toolkit

Semrush’s knowledge base page prices the AI Visibility Toolkit at $99 a month and says it "does not offer a free trial". It includes 25 prompts for prompt tracking, one project, and AI search checks in Site Audit for up to 100 pages. It tracks prompts "on platforms like Google AI Mode and ChatGPT", and the page does not give the full list.

Its strength is that it sits inside Semrush, so a team already paying for SEO data sees AI visibility next to its rankings. The page does not say that the toolkit writes pages.

Agentcites

Agentcites costs $149 a month per website, with up to 50 questions, weekly scans and up to 50 drafted pages a month, as the pricing page shows. It asks ChatGPT, Perplexity, Gemini and DeepSeek today. The first scan of a new website is free and needs no card. It asks 20 of your questions to four assistants, once. It keeps every answer word for word, counts questions that name you apart from those that do not, says why each question was lost, and drafts a page for a loss a page could win.

It is built for depth on fewer assistants rather than breadth. It checks weekly, where four of the other six say on their pages that they check daily. It asks four assistants, not Google AI Overviews, Copilot or Claude. It has no API and no single sign-on. It does not yet tell you whether a page you published changed the answers.

Scrunch

Scrunch’s pricing page lists two plans, read on 11 October 2026. Core is $250 a month for 125 prompts across ChatGPT, Perplexity, Google AI Overviews and Copilot, with a 7-day free trial. Enterprise is priced on request, with a custom number of prompts, nine assistants including Claude, Gemini, Meta AI and Grok, single sign-on and API access.

Scrunch tracks more prompts than Agentcites at its entry price, and asks Google AI Overviews and Copilot, which Agentcites does not. Its approach to fixing is different from writing a page, and it is covered in the section on the fix below.

AthenaHQ

AthenaHQ’s pricing page offers a free Essential tier with 300 credits across five assistants: ChatGPT, Perplexity, AI Overviews, Gemini and Copilot. Starter is $295 a month for 3,600 credits and lists eleven assistants and models by name, including Claude, Grok, DeepSeek, Meta AI and Mistral. The page says "1 credit = 1 AI response". API access is a paid add-on on Starter, and SAML and OIDC sign-on are on Enterprise.

It is the only tool here with a lasting free tier, and its paid plan covers the most named assistants in this list. Starter includes a "Content optimization agent".

Profound

Profound’s pricing page offers a free trial of "50 prompts run daily for 7 days" on ChatGPT, Gemini and Google AI Overviews. Enterprise is priced on request and lists nine engines, including Perplexity, Microsoft Copilot, DeepSeek, Claude and Exa Search, with single sign-on and an API.

Profound covers the widest set of engines on its Enterprise plan and sells the sign-on and API a large company asks for. Its agents page says its agents "generate optimized drafts, and publish directly to your CMS".

Measuring only, or helping you fix it

The tools split into those that measure and suggest, and those that write or change something for you. Which you want depends on whether you have someone to do the writing.

Scrunch’s agent experience page says it detects AI retrieval traffic and routes those visits to an "AI-optimized experience", while "Humans keep the full experience". It also says Google and Bing indexing crawlers still see your normal site. That is a different kind of fix from a new page that people and assistants both read.

How to choose between AI visibility tools

Pick the criterion that matters most to you, and let it decide. No tool here leads on every one.

1. If cost to start is the deciding factor, OtterlyAI at $29 or AthenaHQ’s free tier is the cheapest entry in this list.
2. If you need the most assistants, AthenaHQ’s paid plan and Profound’s Enterprise plan name the most.
3. If you need hundreds of prompts, Peec AI’s largest standard plan and OtterlyAI’s Premium plan track 350 or more, and Scrunch’s Enterprise plan sets its own number.
4. If your team already lives in Semrush, its toolkit keeps everything in one place.
5. If you have nobody to write the pages, look at the tools that write or change something: Agentcites, AthenaHQ, Profound or Scrunch.
6. If you need single sign-on and an API today, Profound, Scrunch, AthenaHQ and Peec AI offer them on their larger plans. Agentcites does not.

Whatever you pick, ask to see the answer behind any figure. A count you cannot open is a count you have to take on trust.

Answers change from run to run. A study by SparkToro and Gumshoe, reported by Search Engine Journal, found the same brand list came back less than 1 per cent of the time across repeated runs. A tool that reports a single ranking position is reporting one draw from a moving set. The method page explains how Agentcites handles that.

What this comparison leaves out

It covers seven tools, read from their own pages on one day. Prices and plans in this category change often, so check each page before you buy.

• It does not test the tools. Nothing here measures whether one tool’s counts are more accurate than another’s.
• It does not include every tool. Ahrefs, HubSpot and others have AI visibility products not listed here.
• It does not weigh reviews, support or ease of use, because none of those can be read from a pricing page.
• It treats "Not stated" as unknown. A tool may do something its pages do not mention.

The comparison pages set Agentcites against five of these tools one at a time, with the same rule about sources.