Track your brand across nine AI engines

Nine: ChatGPT, Claude, Gemini, Perplexity, Grok and Meta Muse, which Rankry asks through their APIs with web search, plus Google AI Overviews and Google AI Mode, which it captures from Google’s live results, and Copilot, which uses Azure GPT with Bing web search through Microsoft’s official API. ChatGPT, Gemini and Perplexity are on every plan; the Google surfaces are included on Pro and available as add-ons on Base; Claude, Grok, Meta Muse and Copilot are add-ons on Base and Pro. For every engine Rankry records whether your brand is named, your position, the sentiment, the competitors named and the sources, and stores the full answer for every new run.

Because the engines disagree. The same buyer question can name your brand in ChatGPT, skip it in Claude and reference a competitor in AI Mode, since each uses a different model, different search and different sources. A blended number hides where you are losing. Rankry keeps every engine separate, and the Model comparison table shows your visibility, position and sentiment for each side by side.

The six assistants are asked through their official APIs: OpenAI and Anthropic with their web search tools, Gemini with Google Search grounding, Perplexity through its Agent API, Grok through xAI’s API with web search, and Meta Muse through Meta’s API with web search. Each prompt is sent as written, once per run, with your location as context where the project has one. Google AI Overviews and AI Mode are captured from Google’s live results through DataForSEO, on desktop, from your project’s location. Copilot uses Azure GPT with Bing web search through Microsoft’s official API.

No. Rankry does not log in to consumer apps, drive a browser or scrape them. It asks through each provider’s API, so answers carry no personal memory, chat history or account settings. That gives a consistent, repeatable read you can compare run to run. For Google AI Overviews and AI Mode, the answer is Google’s own, captured from live search results.

No. They are included on Pro and available on Base as add-ons for $20 a month each, and engines a plan does not include are shown greyed out or blurred in the dashboard. They are also not available for projects in Belarus, Iran or Russia; onboarding warns you if you choose one of those countries.

Yes. When you add prompts, you pick the engines they run on; engines not on your plan are disabled. A CSV import can carry a models column for the same purpose. Prompts without their own selection run on every engine your plan includes. The engines for a prompt are set when it is added.

Reports re-run automatically every day, or every 3 or 7 days if you choose, across the engines each prompt uses. New prompts are answered right away when you add them.

Not in the same way. ChatGPT and Claude return search results and cited links when they search; Gemini returns grounding sources; Perplexity returns its search results; Grok returns only the links it cites inline; Meta Muse and Copilot return the links cited in the answer; AI Overviews and AI Mode carry references in the answer. Web search is not forced for the assistants, so an answer given without a search has no sources.

The basics are shared: be readable by AI crawlers, state clearly what you do, and be described accurately on the sites engines use. The emphasis shifts: sources weigh heavily on Perplexity, Google indexing matters for Gemini, AI Overviews and AI Mode, Bing indexing matters for Copilot, and Grok rewards clear positioning. Each platform page covers what to do there, and Rankry’s Action plan orders the work.

Visibility is the share of an engine’s answers to your prompts that name your brand, with each answer counting once. Position is your average rank when named, share of voice compares your mentions with your tracked competitors, sentiment is judged answer by answer, and Diversity counts how many engines name you. All of it is computed from stored answers, so adding a competitor recalculates history instantly.