AI Tracker. How to track mentions in AI search and analyze brand presence in AI search results - Help Center Topvisor

AI Tracker. Analyze AI mentions

Last modified: 25/05/2026

What you need to track mentions in AI Tracker Going to AI Tracker UI Map Compare Mode

AI Tracker allows you to analyze the presence and mentions of your brand and your competitors in AIs, along with:

Check brand mentions in AI

API Documentation

What you need to track mentions in AI Tracker

Going to AI Tracker

UI Map

  1. Running a check and the Toolbar.
  2. Calendar for picking up checks to build reports for different periods. Works similar to the Calendar in Rank Tracker.
  3. Hide / show Charts, Export and Project Settings.
  4. Charts and Summary.
  5. Searching by prompts.
  6. Mentions / AI snippets mode with full texts of AI responses. Use AI snippets mode to download all texts of AI responses.
  7. Prompts list with their tags. Click on the Prompts column header to sort prompts alphabetically.
  8. Mention check history. — sorting by mentions, — sorting by sentiment. Click on the check date to sort prompts by rank. Hold the cursor over the check date to delete it.
  9. Mention check history table. Tap a cell to see the full AI text answer. What cell icons mean:
    • — brand is mentioned in the AI response, but its position cannot be determined.
    • 2/5 — brand is mentioned in the AI response and occupies the 2nd position out of 5 brands.
    • Dot— green one means the brand is mentioned in a positive tonality, red means the brand is mentioned in a negative tonality. If there is no dot, it means the mention is neutral or the brand is not mentioned at all.
    • — brand is not mentioned in the AI response.

Compare Mode

Select Compare in the AI selection menu to enter comparison mode. Click on a cell with a mention to open the full text of the AI response and switch between different AIs in a single window.

To switch between different checks, click on the arrows. You can switch between different AIs using menu with icons.

Move HTML answer slider to the inactive position to see the response in markdown format.

I ran the check, but as a result I only see the text PROHIBITED_CONTENT. What does it mean?

This message appears in the response from Gemini if the model has considered your prompt potentially unsafe, i.e. related to topics related to adult content, violence, cruelty, etc. Because API security filters are strict and can cause false positives, clarifying or simplifying the prompt may help.