When an AI model answers a question, it can either pull from its built in knowledge or reach for external sources in real time. The difference matters because external sources are something you can actually influence. If a model heavily relies on outside content, your content strategy, PR efforts, and published pages have a real shot at shaping how that model talks about your brand. AI Citations help you see which models lean on external sources the most, so you know where that kind of effort is likely to pay off.
Where to Find AI Citations
You can find the AI Citations section at the bottom of your project's Overview tab. It shows a card for each AI model included in your plan, along with a summary chart below.
How to Read the Data
Each model card displays two values.
The top number represents the total number of external sources that a given AI model cited across all your tracked prompts. This includes repeated citations, so if the same page was referenced in answers to five different prompts, it counts five times.
Pages shows the number of unique pages cited. This tells you how many distinct sources the model actually draws from. For example, Gemini might show 12K total citations but only 902 unique pages, meaning it tends to reuse the same sources frequently.
What Can You Learn From This?
The key insight here is that not all models behave the same way. Some rely heavily on external sources while others lean more on built in knowledge. In the example above, Grok shows 31K total citations while Claude shows almost none. This doesn't mean Claude ignores your brand. It means Claude relies more on its training data and less on real time source retrieval.
This distinction helps you think about your strategy. For source heavy models like Grok, AI Mode, or Gemini, investing in content that can be picked up as an external source is a more direct path to influencing your brand's presence. For models that rely on built in knowledge, the impact of new content is more long term, as it depends on what gets included in future training data.
The bar chart at the bottom gives you a quick visual comparison across all models, making it easy to spot which ones are the heaviest citation users at a glance.
For a deeper look at how your brand ranks across AI models, see our article on Share of Voice. If you have any questions, don't hesitate to contact us.