GEO Audit Tool

AI Visibility & Discovery Audit

Measure how AI models perceive, rank, and recommend your brand — and where competitors have the edge.

Client & context setup

Other names answers might use for the client. Spaces after the commas don't matter.

Use brand names, not website addresses: answers name brands, and only exact names (or aliases) are counted. Add aliases on the same line after a |. Spaces around the | don't matter.

Supply current, verified facts for the grader to compare with each answer. Model-identified contradictions are flagged High risk for human review, not treated as independent proof.

What does hallucination risk mean?

A hallucination is an incorrect or invented claim presented as fact. This rating is an AI grader's review of claims about your brand, not a probability, a full fact-check or a guarantee of accuracy.

  • Low: no obvious issue flagged. Unchecked claims can still be wrong.
  • Medium: specific claims look doubtful and are not covered by supplied facts.
  • High: a contradiction with supplied facts is flagged, or the grader reports a serious factual concern. Review the claim and evidence.
  • Unknown: grading was skipped, unavailable or failed.
  • N/A: your brand was not mentioned, so brand-specific risk was not assessed.

Without brand facts, ratings are model judgement only. The grader can also be wrong, and citations are not automatically verified. Check important figures, dates and claims against authoritative sources before acting.

Query LLMs with questions that specifically mention your brand. Results reveal how AI models perceive, position, and discuss your brand vs competitors.
Query LLMs with generic industry questions that do not mention your brand. Reveals whether you surface naturally in discovery and comparison prompts.

Run a Brand or Unbranded audit to populate the combined dashboard.