Llmrefs

LLMrefs is an analytics tool for what is often called generative engine optimisation, or GEO. As more people ask ChatGPT, Perplexity or Google's AI Overviews instead of clicking through search results, brands want to know whether those answers mention them, where they appear, and which pages get cited. LLMrefs measures this and presents it in a format close to a traditional rank tracker.

How it works

The main design choice is that you track keywords rather than writing individual prompts. You enter a keyword such as a product category, and LLMrefs generates a set of realistic prompts around it, runs them across several AI engines, and combines the results. For each keyword you see brand rankings, share of voice, average position and the number of citations, plus the source URLs the engines relied on. The site shows example views with search volume, a list of the prompts used and a breakdown by engine, along with CSV export and sharing.

Coverage includes ChatGPT and ChatGPT Search, Google AI Overviews, Google AI Mode, Gemini, Perplexity, Claude, Grok and Microsoft Copilot. Results can be geo-targeted across more than 20 countries and 10 languages, which is useful for brands that sell in several markets and see different answers in each.

What you can do with the data

Beyond tracking position, LLMrefs points out content gaps: topics where competitors are mentioned or cited and you are not. Comparing competitors on the same keyword shows who currently owns the AI answer for a topic. The cited sources are often the most actionable part, because they show which articles, reviews or directories the engines trust, and so where outreach or new content is likely to make a difference.

Who it is for

The tool is aimed at in-house marketers, SEO specialists and agencies that already run keyword tracking and want an equivalent for AI search. A free account can be created without a credit card, and paid tiers add more keywords and features. Check the pricing page for current limits.

Limitations to keep in mind

AI answers are not deterministic. The same prompt can produce different brands on different days, so treat the numbers as a sampled trend rather than a fixed rank. Keyword-to-prompt generation is convenient, but it also means you are measuring LLMrefs' idea of how people phrase questions. It is worth reviewing the generated prompts for your key terms to make sure they match how your customers actually ask. Used that way, it gives a practical baseline for a channel that most analytics stacks still ignore.