Cite AI documentation
Cite AI tracks how ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews and Google AI Mode cite your brand in their answers. These guides cover what each feature does, how to read the numbers it produces, and what to change on your site as a result.
Last updated
New here?
Start with the introduction. It walks through creating a project, adding the prompts you want tracked, and reading your first run — about five minutes end to end.
Read the introductionFeature guides
The at-a-glance view: which prompts you appear in, which engines cite you, and how that has moved since the last run.
Share of voice against competitors, citation trends over time, and the gap between where you rank in Google and where you get cited.
Crawls your site and reports the page-level issues that keep content from being retrieved and quoted by answer engines.
Set your brand domain, choose the prompts you want tracked, and add the competitors you are measured against.
Team members and roles, billing and plan limits, notifications, and API access.
Concepts
Background reading on how answer engines pick sources, and what our own measurements say about which changes are worth making.
How a run works
Each tracked prompt is put to every answer engine you have enabled. Cite AI records the full answer, every source the engine cited, and an independent Google SERP snapshot for the same query. Comparing those two is what makes the reports useful: it shows not just whether you were cited, but whether you were cited despite ranking, or missed despite ranking well.
- Runs are scheduled weekly, and your first run starts as soon as you add prompts.
- Results are stored per run, so every number on the dashboard is traceable to a date.
- Competitor domains are scored on exactly the same basis as your own.