Check your site's AI visibility readiness in seconds
Run a free technical preflight for indexability, AI crawler access, schema, core page signals, sitemap discovery, and llms.txt. Then use the manual check below to test what AI assistants actually say.
Free technical preflight
Is your site ready to be found and understood?
Check the public signals AI search systems depend on. No account, API key, or model spend required.
Before you run it
AI visibility checker FAQ
What the free technical check measures, how your website is handled, and how to turn each finding into a safe implementation task.
What does the free AI visibility checker actually test?
It checks eight public technical signals: indexability, AI crawler access in robots.txt, title and description metadata, the primary heading, canonical URL, relevant JSON-LD structured data, question-oriented content, XML sitemap discovery, and an optional llms.txt file. The result is a technical preflight, not a ranking or recommendation guarantee.
Is the readiness score the same as an AI visibility score?
No. The readiness score measures whether a public page exposes technical signals that can help retrieval and understanding. An AI visibility score measures what happens in actual answers: whether the brand is mentioned, how prominently it appears, which competitors appear, and which sources are cited across a controlled question set.
Does the free checker ask ChatGPT about my brand?
No. It reads public website signals without calling ChatGPT, Claude, Gemini, or Grok and without spending model tokens. Measuring current brand mentions requires asking the same neutral buyer questions across providers and storing the answers and citations.
What website data does the checker store?
The free checker does not store the submitted URL or the fetched page contents. It requests only public HTTPS resources needed for the check, applies response-size and timeout limits, and blocks private-network destinations.
Why might AI systems still miss a site that passes?
A passing page can still be limited by CDN firewalls, bot-management rules, JavaScript rendering, provider-specific indexes, weak category authority, unclear brand positioning, or a lack of useful third-party citations. Technical readiness removes avoidable access and interpretation problems; it does not create demand or authority by itself.
What should I do with the “Fix these first” suggestions?
Start with the highest-scoring gap, confirm the finding in your actual code or hosting configuration, and make the smallest scoped correction. Each suggestion can be copied individually, or you can use “Copy for your agent” to create a ready-to-paste implementation prompt with the URL, evidence, safeguards, and verification requirements.
How can I check my brand's actual AI visibility for free?
Write ten neutral buyer questions that do not name your brand, ask each in a fresh session across the assistants your buyers use, and record mentions, prominence, competitors, and cited domains. This is a useful directional sample, but personalization and small sample size mean it should not be treated as a repeatable benchmark.
How often should I rerun the checker?
Rerun it after changing metadata, robots rules, structured data, sitemaps, page structure, or hosting controls. For actual AI visibility, wait long enough for providers to recrawl and refresh their indexes, then repeat the same frozen question set so the comparison remains meaningful.
The actual-answer check
Technical readiness is not the same as being recommended
A technically perfect site can still be absent from AI answers. Use this five-step manual check to sample real mentions while keeping its limitations in view.
- 01
Write ten neutral buyer questions
Phrase them the way a buyer who has never heard of you would: “best X for Y,” “how do I solve Z.” Do not include your brand name—discovery questions test whether AI finds you on merit.
- 02
Ask them in fresh sessions
Use a new chat for every question, with web search enabled where the assistant supports it. Repeat the same questions across the assistants your buyers actually use.
- 03
Log mentions, position, and sources
For each answer record whether you were named, whether you led the answer or trailed it, which competitors appeared, and which domains were cited.
- 04
Ask a direct diagnostic question
Ask each assistant what it knows about your brand by name. Wrong facts, stale descriptions, or confusion with a similarly named company are findings in themselves.
- 05
Total it into a rough score
Mentions divided by answers gives a crude visibility rate. Note it with the date—models change, so a check without a timestamp is unrepeatable.
Where manual checks break
Four reasons the quick check misleads
Personalization and memory
Sample size
No stored evidence
Moving targets
What a rigorous check measures
Four separate metrics beat one mystery score
Visibility
How often your brand appears in neutral discovery answers.
Prominence
Whether you lead the answer, make a shortlist, or appear incidentally.
Share of voice
Your answer-level mentions compared with selected competitors.
Citations
How often eligible answers cite your claimed domain as a source.
Coverage and sentiment complete the picture: how many planned answers actually returned with valid web grounding, and how the models talk about you when you do appear. The AI visibility guide defines each metric in depth.
The structured version
What the $9 audit automates
100 Questions runs the same check with controls a manual pass cannot match: a frozen set of 25 questions—20 neutral discovery questions plus 5 direct diagnostics—sent identically to OpenAI, Claude, Gemini, and Grok with shared web-search grounding, for 100 planned answers per run.
Answers without valid web sources are excluded from score denominators instead of quietly inflating them, every metric links back to stored answers and citations, and a rerun reuses the identical questions so before-and-after comparisons mean something. The output is a client-ready report with five evidence-linked actions—see the complete sample report.
Run the real check