Skip to content
Back to writing
Method

LLM visibility: how to check whether AI assistants recommend your brand

LLM visibility is whether ChatGPT, Gemini, Perplexity and Google's AI answers name your brand when buyers ask about your category. Here is how to check it properly.

5 Nov 2026 - 5 min read
On this pageShow 8Hide

Ask ChatGPT for the best option in your category and read the answer. Then ask Gemini and Perplexity the same thing, and search it on Google to see whether an AI Overview appears. Whatever those answers say is your LLM visibility for that one question.

It is easy to look once, at one prompt, in one assistant, and draw a conclusion from it. This guide covers how to check properly: by hand first, then with software, and what to do with what you find.

What LLM visibility means in practice

LLM visibility is whether AI assistants name your brand when a buyer asks about your category, and whether they cite your pages when they do. It has three parts worth separating:

  • Presence: is your brand in the answer at all?
  • Position: is it the first recommendation, or the fifth in a list?
  • Source: does the answer link to your site, or does it describe you from what other sites say?

Each one can move without the others. A brand can be named in most answers and cited in almost none, for example when the engines describe it from reviews and roundups rather than its own pages.

Free scan

Check your AI visibility free

See which buying questions ChatGPT, Perplexity, Gemini and Google's AI answers name you for, and who they name instead. No call, no card.

The manual method

You can do a useful check in an hour with nothing but the assistants themselves.

1. Write five buyer prompts

Pick one topic you sell into and write five prompts the way a buyer would type them. Cover different angles: the plain category ("best project management software for agencies"), a specific need, a specific kind of buyer, the outcome they want, and "alternatives to" a competitor.

2. Run them in each engine

Turn off memory and chat history where you can, so your own history does not shape the answer. OpenAI says that with Reference chat history on, ChatGPT can use past conversations to personalize responses. Run each prompt in ChatGPT, Gemini and Perplexity, and search it on Google to see whether an AI Overview appears for it.

3. Record what came back

For every answer, note the brands named, the order, whether you are one of them, and the pages cited as sources where the engine shows them. A simple sheet with one row per answer is enough.

4. Read the sources

The sources are the most useful part. Open the pages the engines cited when a competitor was named and you were not. Those pages are what the engines are reading about your category. If you are not on them, that explains a lot of what you just saw.

Where the manual method breaks

The manual check is worth doing once, because it teaches you what the numbers mean. It does not scale:

  • Answers vary. The models are not deterministic: OpenAI's API documentation says outputs may differ from request to request, and the Gemini app lets you regenerate an answer and flip between versions. One reading is a sample.
  • The work multiplies. Five prompts on four engines is 20 answers for one topic. Ten topics is 200, and you want them more than once.
  • History is hard to keep. A sheet tells you what happened today. It does not tell you that a competitor appeared in Perplexity last Tuesday and has been there since.

The software method

Software runs the same prompts on a schedule, stores every answer and its sources, and shows the trend. When you compare tools, check four things:

  1. Which engines it reads. ChatGPT, Gemini, Perplexity and Google AI Overviews are a sensible starting set. Check that the ones your buyers use are covered on the plan you would pay for.
  2. How often it checks. Daily readings show you what changed and when. Weekly is slower to show cause and effect.
  3. Whether it keeps the answers. A score with no answer text behind it is hard to act on. You want the full answer and the pages it cited.
  4. What it costs per prompt. Most plans are priced by the number of prompts, or by credits spent per answer, so compare on that basis.

Our roundup of AI search visibility tools compares the main options on exactly those points, with US prices taken from each vendor's own site.

Run a free check now

We built the LLM visibility checker to do the manual method for you, once, for free. You enter your domain, confirm five buyer prompts it drafts from your site, and about two minutes later you get 20 answers: each prompt on Google AI Overviews, ChatGPT, Gemini and Perplexity. You see the full text of each answer, the pages behind it, the brands named instead of you and where your site ranks on Google for the keyword behind the prompts. No card and no email.

If you want the same check every day, that is the alwaystracked plan: 10 clusters, which is 50 prompts and 10 keywords, checked daily, from $129 a month with a 14 day free trial.

What a good result looks like

There is no universal score that means good, but there are clear signs:

  • You are named on the category prompt in most engines, not just one.
  • You are named early in the answer, not at the end of a long list.
  • At least some answers cite your own pages, which means the engine is reading you directly.
  • You appear on the comparison prompt for the competitor you lose most deals to.

A weak result is just as clear: missing from the category prompt everywhere, named only on prompts that include your brand name, and never cited.

What to do with a weak result

Work from the sources outward. If the cited pages are roundups that leave you out, getting onto those pages is the most direct fix. If the engines cite competitors' own pages, look at what those pages answer that yours do not. And if you show up in Google's results but not in the assistants, the guide to how to rank in ChatGPT covers what moves that engine specifically. If the engines struggle to read your own pages, our guide to LLM SEO covers the on site fixes, and once you are tracking, AI share of voice turns your readings into a number you can compare with competitors.

Questions

What is LLM visibility?
It is how often large language model assistants, such as ChatGPT, Gemini and Perplexity, and Google's AI Overviews name or cite your brand when someone asks about the category you sell in.
Can I check LLM visibility for free?
Yes. You can run prompts by hand in each assistant and record the answers, or use a free checker that runs a set of buyer prompts across several engines and shows you every answer.
Why does my brand show up in one assistant but not another?
They are different products with different models and search systems. Even within Google, AI Overviews and AI Mode may use different models, and Google says the responses and links they show will vary. That is why a check should cover each engine separately rather than one.
How is LLM visibility different from SEO rankings?
A ranking is a position in a list of links. LLM visibility is whether you appear in a written answer at all, and whether the answer links to you. Google says meeting its requirements does not guarantee a page will be shown, and AI answers can cite pages outside the top ten results, so the two need measuring separately.
Free scan

Check your AI visibility free

See which buying questions ChatGPT, Perplexity, Gemini and Google's AI answers name you for, and who they name instead. No call, no card.