Skip to content
RookAtlas

RookAtlas facts

Facts, status and coverage

RookAtlas is an AI visibility tracker: it measures how often AI answers name a brand on eight engines, with a 95% interval on visibility and citation rates, and turns the gaps into evidence-graded changes a team approves.

Last updated 12 November 2026. A reference for reviewers, journalists and anyone comparing AI visibility tools; the method has the detail.

Status

RookAtlas opens soon. Sign-up opens with live measurement, after our provider pilot.

On 12 November 2026, the first live runs asked 100 Turkish and 100 English prompts on each of ChatGPT, Gemini, Google AI Mode and Google AI Overviews, and all 800 collection tasks completed with no failures; Perplexity, Claude, Grok and Muse are still to be tested.

RookAtlas is a product of RookAtlas LLC, a Wyoming limited liability company. More on the publisher and our editorial policy is on About RookAtlas.

Engines and how each is asked

ChatGPT, Gemini, Perplexity, Google AI Mode and Google AI Overviews are read logged out from their own interfaces, through contracted data providers. Claude, Grok and Muse are asked through their makers' official APIs with web search enabled, so their answers can differ from the apps'. Every engine's answers combine into one visibility number, each engine also gets its own, and users choose which engines count.

ChatGPT
RookAtlas collects ChatGPT's answers from the app's own interface, logged out, the way a new visitor sees them, and keeps every answer whole with its sources.
Gemini
RookAtlas collects Gemini's answers from the Gemini app's own interface, logged out, and keeps each answer whole with the sources it showed.
Perplexity
RookAtlas collects Perplexity's answers from its own interface, logged out, and keeps every numbered source with the answer.
Google AI Overviews
RookAtlas reads Google's results page with the AI Overview loaded. A search that shows no Overview is recorded as such and left out of your visibility, so a missing Overview never counts as being ignored.
Google AI Mode
RookAtlas collects AI Mode answers from Google, logged out, and keeps each answer whole with the pages it linked.
Claude
RookAtlas asks Claude each prompt with web search enabled, at most one search per answer, and keeps the answer whole with its citations.
Grok
RookAtlas asks Grok each prompt with web search enabled and X search off, so its searches cover the open web, and keeps their citations.
Muse
RookAtlas asks Muse's model each prompt with web search enabled and keeps the answer whole with its citations.

A paid prompt run, and each of Free's 10 sign-up credits, asks all eight. Free's monthly credits run on ChatGPT, Gemini, and AI Mode.

What it measures

  • How often answers name the brand (mention rate), where they place it, its share of voice against named competitors, and the sentiment of each mention.
  • Which pages and domains answers cite, the brand's own citation rate, and the searches engines report running.
  • A 95% interval on visibility and citation rates, clustered by prompt, with low-confidence and too-little-data labels when answers or prompts are few.
  • Versioned prompts, panels and competitor sets: an edit starts a new version, and past numbers keep the version they were measured on.
  • Every answer kept whole with its sources: open in the app, as a CSV download, and through the API and MCP server.

Definitions are in the glossary and, with their formulas, in the method.

Plans and credits, planned at launch

One credit runs one prompt in one market on all eight engines. Every run that spends credits shows its price before it starts.

Free
$0. 10 credits at sign-up, then 5 a month, 3 brands, 1 seats.
Starter
$45 a month, $36 billed annually. 150 credits a month, 10 brands, 2 seats.
Pro
$79 a month, $63 billed annually. 500 credits a month, 100 brands, 5 seats.
Agency
$129 a month, $103 billed annually. 1,000 credits a month, 500 brands, unlimited seats.
Scale
$259 a month, $207 billed annually. 2,000 credits a month, 1,000 brands, unlimited seats.

Top-ups: 50 credits for $15, 100 credits for $28, and 200 credits for $50, valid for 12 months. Full detail on pricing.

After a gap is found

Prompt gaps, source gaps and an AI-access audit, each with an evidence grade (official, experiment, correlation, vendor or weak); sourced drafts a person approves; outreach to cited sites in three-step sequences; white-label reports and share links from Pro.

REST API and MCP server
Built, on every plan, Free included.
Signed webhooks
Built.
Slack and email alerts
Built; Slack from Starter.
WordPress publishing with IndexNow
Built; tested on recorded responses so far.
Bing Webmaster Tools AI Performance
Built, as an export you upload.
Gmail for outreach
Built; sending for customers waits for Google's verification of the connection.
Google Search Console and Google Analytics 4
Coming, after the same Google verification.
Client portal on your own domain
Coming.

What RookAtlas doesn't do

  • Microsoft Copilot is not currently measured; Bing's AI Performance export can be uploaded instead.
  • RookAtlas does not buy placements in AI answers or sponsored articles; its outreach is for earned mentions.
  • It doesn't serve AI crawlers a different version of a customer's pages: the same public content serves people and crawlers.
  • Drafts are published and outreach emails are sent only after a person approves them.
  • Engines are asked without a signed-in account or personal history, so answers don't reflect any one user's chats.
  • As of 12 November 2026, the site shows no customer logos, ratings or case studies.

Compared with other tools

Side-by-side pages, every fact about another tool sourced to its own pages and dated: all comparisons, Peec AI, Otterly.AI, Surfer, AthenaHQ, Profound and Scrunch.