last round today 18:25 UTC 2163 live 12 in the graveyard next round in —
TrueTier
AI search · free tool · no signup

AI visibility checker

Can ChatGPT, Claude, Gemini and Perplexity read your site — and what do they answer when a customer asks about what you sell? Part 1 checks access, readability and how clearly the site says who you are, in about 20 seconds, free. Part 2 asks real AI engines your customers' questions and shows every answer.

10 of 10 free runs left this hour. Used them up? An extra run is $1.49 and works for every tool here, no account needed.

How it works

  1. Part 1 reads your robots.txt for 15 AI crawlers and requests the page as several of them, to catch firewalls that answer 403 or a challenge page.
  2. It measures how much text is in the raw HTML (most AI crawlers do not run JavaScript), checks headings, title, language and canonical, and looks for structured data that says who you are.
  3. Four sub-scores with their weights add up to the 0–100 score; every point comes from a check you can see. The fixes (JSON-LD built from your own page, robots.txt lines) are ready to paste.
  4. Part 2 (account): a model reads your home page and proposes 8 questions a customer would ask, in your site's language — you can edit them. Each question goes to each engine twice, with web search on.
  5. You get, per engine: how often you are named, how often your own site is cited, your place in lists, who is named instead, the sources the AI cites, claims that contradict your site — and every answer in full, with an action plan ordered by benefit.

Questions people ask

Which AI engines does the answer test use?

Claude (Anthropic) with its web search tool always; Gemini with Google Search grounding and OpenAI's model with web search when they are switched on. The report names exactly which engines and models answered. We use their APIs — the consumer apps (ChatGPT, the Gemini app, Google AI Overviews) can answer differently.

Why does each question run twice?

Because AI answers vary from run to run. Two runs per question show whether a mention is stable or luck. Treat the result as a sample, not a ranking.

Do you show the questions and the answers?

Yes, all of them, word for word, with the sources each answer cited. You can check every number in the report against the answers it came from.

Does llms.txt help?

We show whether you have one, with zero weight. In May 2026 Ahrefs reported that 97 % of llms.txt files were never fetched, and Google said AI-specific files get no special treatment. Access, readable text and clear facts matter more.

Will this get me more customers?

We do not promise that. The test shows where you stand today and what is missing; the engines decide what they cite. Re-test after 30 days to see whether the changes moved anything.

Should I block AI training crawlers?

That is your choice. Blocking training crawlers (GPTBot, ClaudeBot, Google-Extended, CCBot…) does not remove you from AI search answers; blocking the search and user crawlers (OAI-SearchBot, ChatGPT-User, Claude-SearchBot, PerplexityBot…) does.

Free, no signup. Runs on our servers, nothing is installed. Want all of it at once? The full site check runs 30+ tests in one go.