Do any AI visibility tools track Meta AI?

Most AI visibility tools track some mix of ChatGPT, Perplexity, Gemini, Claude and Google's AI surfaces. Meta AI is usually missing. Our August 2026 evidence review, which checked claims against primary sources, found no vendor tracking Meta AI at all. 99Visibility is no exception: its paid plans audit ChatGPT, Google AI Overviews, Gemini, Perplexity and Claude.

The gap is odd because Meta has said more than anyone else about how to get in. Meta's crawler documentation states: "Allowing Meta-WebIndexer in your robots.txt file helps us cite and link to your content in Meta AI's responses." No other operator in the evidence makes a first-party statement tying crawler access to citation.

What that means for you:

• If a tool claims Meta AI coverage, ask how it collects answers and whether it shows the links Meta AI returned.
• If no tool covers it, you still control the access step, which is the part Meta describes.
• Track answers manually until coverage exists. Our page on tracking Meta AI answers about your brand sets out a method.

Meta's sentence is an operator assertion, not a measured result. It tells you the mechanism: Meta-WebIndexer feeds the index that Meta AI cites from. It does not tell you how much allowing it helps, how often Meta AI searches at all, or how it picks between sources.

That fits the wider evidence. A July 2026 critical survey of 45 GEO studies concluded that no reviewed technique shows a stable, longitudinal, cross-platform causal effect on discoverability. The levers that hold up are access levers. They work by removing blocks, not by adding anything.

Keep the four stages apart:

• Parametric: what Meta's model knows with no web access. Nobody controls this in the short term.
• Access: whether Meta-WebIndexer may fetch and store your page. You control this.
• Retrieval: whether your page enters the answer for a given question. Meta controls this.
• Generation: whether a retrieved page gets cited. The model controls this.

Allowing the crawler fixes the second stage only. It makes you eligible. It does not promise a citation.

Robots.txt rules are per bot, and Meta runs several. Mixing them up is the most common mistake.

The same split exists at OpenAI, where GPTBot governs training and OAI-SearchBot governs ChatGPT search, and OpenAI says each setting is independent. A blanket "block all AI" rule can remove search eligibility you wanted to keep. If you want to stay out of training but stay citable, you write separate rules for each bot.

These checks need no tool subscription. Run them in order.

1. Read your robots.txt. Look for any rule that names Meta-WebIndexer or a wildcard Disallow that catches it. If you want Meta AI to cite you, an explicit allow is clear:

User-agent: Meta-WebIndexer
Allow: /

2. Test the firewall, not just the file. A CDN or firewall can block a crawler that robots.txt allows. Request a key page with Meta-WebIndexer's user-agent string and check the response. A 200 passes. A 403, a 429 or a challenge page means the crawler is turned away, and nothing in your analytics will tell you.
3. Check your CDN's AI crawler settings. Since 2026-09-15, Cloudflare blocks Training and Agent crawler categories by default on ad-displaying pages for new domains, new sites on existing accounts and free-tier customers. Confirm which category your provider puts each Meta crawler in.
4. Check rendering. Vercel's crawler study found ChatGPT and Claude crawlers fetch JavaScript but do not execute it, and a June 2026 retest found the same for Meta-ExternalAgent. Nothing in the evidence shows Meta-WebIndexer behaves differently. Put prices, product names and key facts in the raw HTML.
5. Check snippet controls. A stray noindex, nosnippet or data-nosnippet inherited from a template can suppress content you meant to show.

Access checks tell you whether Meta can reach you. They do not tell you what Meta AI says. For that, you need answers.

Say you run a hypothetical bookkeeping firm in Leeds. Write five to ten buyer questions, such as "who does bookkeeping for small cafes in Leeds". Ask each one in Meta AI on a fixed schedule. Log the date, the full answer, whether your brand appears, which links appear, and any wrong facts.

Rules that keep the log useful:

• Use the same wording every time, so changes reflect Meta AI, not your prompts.
• Record whether the answer showed links. An answer with no links may come from the model's memory, not live search.
• Flag wrong prices, names or locations separately from absence. Being misdescribed is a different problem from being missing, covered in our page on detecting brand misinformation in AI answers.
• Treat counts as a sample. Any prompt set is a proxy for the queries the engine actually runs.

The same discipline applies across engines. See ChatGPT prompt tracking for a reproducible version of this method.

If your buyers use Meta's apps, Meta AI coverage is a fair question to put to any vendor. Ask four things:

• Does it query Meta AI directly, and how?
• Does it report Meta AI separately, or blend it into one score?
• Does it show the links Meta AI returned, not just mentions?
• Does it check whether Meta-WebIndexer can actually reach your site?

A single blended score hides the engine that matters to you, which is why one dashboard for every AI engine is harder than it sounds. For the wider picture, read how AI visibility tracking works. Until coverage arrives, pair a tool for the engines it does track with your own Meta AI log and access checks.