Which AI Engines Do AI-Visibility Tools Track? The 2026 Census

The 27 August 2026 directory record contained far more ChatGPT tags than Microsoft Copilot tags. That historical arithmetic is recoverable, but it is not proof that every tagged product could monitor the named consumer engine on the plan a buyer would purchase. This correction preserves the historical counts and full table while narrowing them to what the recovered evidence actually measures.

Editorial correction, 5 September 2026: recovered the original-sized population and engine-tag arithmetic from warehouse collector run 221171; corrected claims of universal live reproducibility, vendor absence and plan entitlement; and added the API/search-surface distinctions missing from the original interpretation. This is not a new engine benchmark or whole-market re-verification. Original publication, creation and verification clocks remain unchanged.

Editorial correction, 12 September 2026: ModularDot’s listing was withdrawn from the index on 12 September 2026. Its name in this report therefore appears as plain text rather than a link. The figures in this report are unchanged: they describe a dated reading taken at collector run 221171 on 27 August 2026, when the listing was published.

Corpus and method

102 Published listings were captured in CitedIndex warehouse collector run 221171.

86 of those captured listings remain after excluding the 16 carrying the citation-services category in that same run.

85 listings remain in the historical engine table after also excluding AI Visibility Report Group.

The collector observed the directory at 2026-08-26 22:00:03 UTC, labelled 27 August in Europe/Amsterdam. It captured membership and taxonomy under the same run identifier. The original exclusion of AI Visibility Report Group reflected unresolved engine names, not a proven zero. The original table retained GEOrank and ModularDot (then Nostalgai) as non-monitoring products. The agency-category exclusion defines this software comparison; it does not claim that agencies cannot cover particular engines.

The six columns are ChatGPT, Perplexity, Gemini, Claude, Microsoft Copilot and Google AI Overviews. Other products and surfaces such as Google AI Mode, Grok and DeepSeek are outside these columns, not counted as unsupported. Later additions and changed tags are not backdated into this population.

Limitations before interpretation

The recovered taxonomy reproduces the table, but a directory tag is an editorial classification, not an authenticated monitoring result. The old registered queries used live tags while their sentences referred to earlier editions. A later tag reconciliation also changed this directory’s classifications. Rerunning today’s query cannot verify an August count, and a missing tag is not a vendor denial.

The historical tags do not distinguish consumer UI collection from provider API sampling, enabled search from observed search, synthetic prompts from real user queries, or plan-level availability. We have not recovered contemporaneous raw vendor pages establishing each of those properties for every row. The old unqualified claims that the tagged products “track” each consumer engine are withdrawn to that extent. The arithmetic below is retained as a dated record of the directory’s coverage classifications, not a corrected current vendor-support census.

1. The recorded engine distribution

83 of the 85 historical table listings carried the ChatGPT tag in collector run 221171.

75 of the 85 historical table listings carried the Perplexity tag in collector run 221171.

73 of the 85 historical table listings carried the Gemini tag in collector run 221171.

62 of the 85 historical table listings carried the AI Overviews tag in collector run 221171.

59 of the 85 historical table listings carried the Claude tag in collector run 221171.

34 of the 85 historical table listings carried the Copilot tag in collector run 221171.

Recorded engine tag	Listings	Share of the 85
ChatGPT	83	98%
Perplexity	75	88%
Gemini	73	86%
AI Overviews	62	73%
Claude	59	69%
Copilot	34	40%

Copilot was the least common of these tags. That supports a narrower conclusion than “nearly everyone skips Copilot”: the directory recorded it less often, with both absent support and unresolved documentation capable of producing an untagged row.

Method: count each named tag once per included listing in run 221171; divide by the fixed table population and round percentages to whole numbers. No model response was generated for this correction.

2. Broad recorded coverage still had a common gap

28 of the 85 historical table listings carried exactly five of the six engine tags in collector run 221171.

17 of those 28 five-tag listings lacked the Copilot tag in collector run 221171.

7 of those 28 five-tag listings lacked the Claude tag in collector run 221171.

4 of those 28 five-tag listings lacked the AI Overviews tag in collector run 221171.

The missing-Copilot group was Archytas AISpy, Cision AI Visibility Dashboard, CiteMentor, Directree GEO Monitor, Frase, Indexly, Keyword.com AI Visibility, Knowatoa, Local Dominator AI Tracker, Nightwatch, Orchly AI, Rankfender, Rank Prompt, Searchable, SEORCE, Trakkr and WhyIQ AI Radar. The missing-Claude group was Ahrefs Brand Radar, Bloomiro, Dageno AI, Gauge, Peec AI, RadarKit and WebGlazer. Addlly AI, Adobe Brand Visibility, Analyze AI and Meltwater GenAI Lens lacked the AI Overviews tag.

These are historical tag gaps, not an instruction to reject those products today. A plan allowing a customer to choose a subset of a broader supported engine menu is also not the same thing as a product supporting only that subset. The earlier Addlly discussion conflated those concepts; a plan ceiling must not be used to manufacture engine-specific negatives.

Method: sum the six binary tag indicators per listing, select totals of five, then inspect the absent tag in that same capture.

3. The all-six and lower-coverage groups

22 of the 85 historical table listings carried all six engine tags in collector run 221171.

50 of the 85 historical table listings carried at least five engine tags in collector run 221171.

17 of the 85 historical table listings carried exactly four engine tags in collector run 221171.

14 of the 85 historical table listings carried exactly three engine tags in collector run 221171.

2 of the 85 historical table listings carried exactly two engine tags in collector run 221171.

2 of the 85 historical table listings carried none of the six engine tags in collector run 221171.

83 of the 85 historical table listings carried at least one of the six engine tags in collector run 221171.

The all-six group was AIclicks, AI Sightline, ApexGEO, Brandlight, Citadex, Cognizo, Conductor, EdenRank, Elmo, Evertune, Goodie AI, LLM Pulse, LLMrefs, Mentionable, Omnia, Profound, Promptwatch, Qwairy, Rankability, RankScale, Rankshift and Writesonic. GEOrank and ModularDot were the all-untagged rows; the original research described page-diagnostic/readiness tooling rather than engine monitors.

Method: group the fixed population by the sum of its six tag indicators. The table below preserves the individual assignments, rather than replacing this historical distribution with the size of today’s directory.

What engine coverage does and does not buy

Plan entitlement is a separate question. Profound’s pricing page read for this correction lists Starter as ChatGPT-only, Growth with three answer engines, and Enterprise with up to nine. The old claim that everything beyond ChatGPT requires Enterprise is not a correct current reading: Growth exists between them. These current plan statements are not used to rewrite the August tag counts, and a numerical “up to” allowance is not proof that every named engine is included.

A provider API call is not necessarily a consumer-engine observation. Canonry’s README documents provider credentials for its measurements. A provider model and a consumer chat product can differ in search settings, context and retrieval. An engine label alone does not tell a buyer which surface was sampled.

Requested search is not observed search. OpenAI’s web-search documentation says the model can choose whether to search when the tool is available. Its output represents actual use with a web_search_call and separately carries URL-citation annotations. A request that enables search, a completed search action and a citation in the final answer are different observations. This report did not inspect response traces, so its tags establish none of them.

A readiness score is not a citation measurement. A structural audit, a simulated buyer prompt and a retrieved answer can all be useful, but they answer different questions. For example, AEOlens’ pricing descriptions separate structural scores, AI crawler analytics and AI Buyer Simulations. None should be inferred from another merely because the page carries a model logo.

Use the comparison table and six-engine collection as navigation, then inspect each listing’s own sources and dates. Those live pages are not immutable outputs of this historical capture. Match the observed surface and plan to the audience you need to measure, not only the largest engine count.

The full breakdown

The preserved August table, sorted by recorded tag count. “Yes” means the tag was recorded in the recovered capture; “No” means the tag was absent, not that a vendor denied support. The two all-No rows were originally classified as non-monitors. Product names and links retain subsequent naming corrections, including ModularDot.

Tool	Category	ChatGPT	Perplexity	Gemini	Claude	Copilot	AI Overviews	Recorded tags
AI Sightline	Rank tracking/Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
AIclicks	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
ApexGEO	Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Brandlight	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Citadex	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Cognizo	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Conductor	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
EdenRank	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Elmo	Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Evertune	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Goodie AI	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
LLM Pulse	Rank tracking/Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
LLMrefs	Rank tracking/Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Mentionable	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Omnia	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Profound	Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Promptwatch	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Qwairy	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Rankability	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
RankScale	GEO optim./Rank tracking/Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Rankshift	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Writesonic	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	Yes	6
Addlly AI	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	No	5
Adobe Brand Visibility	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	No	5
Ahrefs Brand Radar	Visibility	Yes	Yes	Yes	No	Yes	Yes	5
Analyze AI	GEO optim./Visibility	Yes	Yes	Yes	Yes	Yes	No	5
Archytas AISpy	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Bloomiro	Visibility	Yes	Yes	Yes	No	Yes	Yes	5
Cision AI Visibility Dashboard	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
CiteMentor	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Dageno AI	GEO optim./Visibility	Yes	Yes	Yes	No	Yes	Yes	5
Directree GEO Monitor	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Frase	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Gauge	GEO optim./Visibility	Yes	Yes	Yes	No	Yes	Yes	5
Indexly	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Keyword.com AI Visibility	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Knowatoa	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Local Dominator AI Tracker	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Meltwater GenAI Lens	Visibility	Yes	Yes	Yes	Yes	Yes	No	5
Nightwatch	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Orchly AI	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Peec AI	Rank tracking/Visibility	Yes	Yes	Yes	No	Yes	Yes	5
RadarKit	GEO optim./Visibility	Yes	Yes	Yes	No	Yes	Yes	5
Rank Prompt	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Rankfender	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Searchable	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	Yes	5
SEORCE	GEO optim./Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	Yes	5
Trakkr	Visibility	Yes	Yes	Yes	Yes	No	Yes	5
WebGlazer	Visibility	Yes	Yes	Yes	No	Yes	Yes	5
WhyIQ AI Radar	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	Yes	5
AEOlens	GEO optim./Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	No	4
auto-geo	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
BabyLoveGrowth	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
Canonry	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
CheckThat.ai	Visibility	Yes	Yes	No	Yes	No	Yes	4
GEO Tool	GEO optim./Visibility	Yes	Yes	No	Yes	No	Yes	4
GeoArk AI	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
Mangools AI Search Watcher	Rank tracking/Visibility	Yes	No	Yes	Yes	No	Yes	4
Otterly AI	Rank tracking/Visibility	Yes	Yes	No	No	Yes	Yes	4
Prizvox	Visibility	Yes	Yes	Yes	Yes	No	No	4
RankLens	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	No	4
Scrunch AI	Visibility	Yes	Yes	No	Yes	No	Yes	4
Semrush AI Visibility Toolkit	GEO optim./Rank tracking	Yes	Yes	Yes	No	No	Yes	4
Similarweb AI Search Intelligence	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
Surfer SEO	GEO optim./Visibility	Yes	Yes	Yes	No	No	Yes	4
TrueRanker	Rank tracking/Visibility	Yes	Yes	Yes	Yes	No	No	4
VisibAI	GEO optim./Visibility	Yes	Yes	Yes	Yes	No	No	4
AthenaHQ	Visibility	Yes	Yes	No	No	No	Yes	3
CiteLens	GEO optim./Visibility	Yes	Yes	No	Yes	No	No	3
Distribb	GEO optim.	Yes	Yes	Yes	No	No	No	3
Emergine	Rank tracking/Visibility	Yes	No	Yes	No	No	Yes	3
GeoRankers	Visibility	Yes	Yes	Yes	No	No	No	3
geoSurge	GEO optim./Visibility	Yes	No	Yes	Yes	No	No	3
HubSpot AEO	GEO optim./Visibility	Yes	Yes	Yes	No	No	No	3
Local Falcon	Rank tracking/Visibility	Yes	No	Yes	No	No	Yes	3
SE Visible	Rank tracking/Visibility	Yes	Yes	No	No	No	Yes	3
Sleepwalker	GEO optim./Visibility	Yes	Yes	Yes	No	No	No	3
Syntropic AI Visibility Audit	Visibility	Yes	No	Yes	No	No	Yes	3
Ubersuggest	GEO optim./Visibility	Yes	No	Yes	No	No	Yes	3
Vismore	GEO optim./Visibility	Yes	Yes	Yes	No	No	No	3
ZipTie	GEO optim./Visibility	Yes	Yes	No	No	No	Yes	3
Amplitude AI Visibility	GEO optim./Visibility	Yes	No	No	No	No	Yes	2
Morningscore	GEO optim./Rank tracking/Visibility	Yes	No	No	No	No	Yes	2
GEOrank	GEO optim.	No	No	No	No	No	No	0
ModularDot	GEO optim.	No	No	No	No	No	No	0

Historical changes and correction record

The earlier revision notes reported changes on 12, 14, 16, 17, 20, 24, 25 and 27 August 2026, including the addition of GeoArk AI and Nostalgai (renamed ModularDot on 28 August), changes to Evertune’s record, and withdrawn Copilot assignments for Nightwatch and SEORCE. Those notes remain editorial history; this correction recovered run 221171, not the raw execution records behind every earlier pass. It therefore withdraws the absolute assertion that all earlier rows were independently rechecked and unchanged, rather than claiming to have repeated those past checks.

Two earlier corrections remain important: the 14 August gap-comparison sentence was arithmetically wrong and was withdrawn, and the 17 August Addlly note reported different content in a rendering browser than a static fetch. Neither a later corrected reading nor a changed tag establishes when a vendor added support. We do not describe those editorial discoveries as product launches.

Reproducing and citing the retained findings

The retained counts use the published CitedIndex listing records and their category and engine-tag assignments from archived observation 221171. Match records from that same observation by listing identity. Exclude the captured citation-services category and, for engine counts, AI Visibility Report Group. Count each listing once in each of the six named engine columns. This identifies the recovered historical observation, not today’s directory population or a separately preserved set of vendor-page captures.

Cite these as the directory’s recorded August engine classifications, including the limitation that tags do not prove consumer-surface monitoring or current support. The unqualified vendor-support interpretation and claims of testing each product are not retained findings. For the analogous interface distinction, see the MCP report and API report.