The category leaderboard for AI search.

Illustrative category index — unaided mention rate, average position, citation rate and surface coverage. Not the five-part score from the AI Visibility Tracker.

An AI Visibility Index is a category-level score of how often — and how high — AI engines recommend your brand when someone asks about your market without naming you. It is not a Google ranking and it is not a one-off URL check: it is a leaderboard of unaided brand awareness across ChatGPT, Gemini, Claude, Perplexity, Grok, DeepSeek, Google AI Overviews and AI Mode, refreshed on a schedule so you can see the trend.

The tracker measures the prompts you choose — your questions, your competitors, the literal answer text stored every week. The index measures the category: unaided recommendations, average position, who leads the industry leaderboard, which sources models cite, and which brands co-occur. Use the index to see where you sit in the market; open the tracker when you need the exact answer that put a rival ahead of you.

The SEO Visibility Index is the official Sistrix score for organic Google visibility of your project domain. The AI Visibility Index is about LLM and AI-search recommendations. Two different “index” products, two different SERPs — they are not translations of each other and they are not interchangeable metrics.

The free AI Visibility Checker is a one-off audit of a single URL: no category leaderboard, no history, no account. The AI Visibility Index is continuous category measurement inside a TrueRanker project — weekly scans, rank, score, sources and co-occurrence over time.

In the rank tracker, AI Overviews and AI Mode are columns on a keyword SERP: whether Google showed an AI answer for that query and whether it named or linked you. In the AI Visibility Index they are category surfaces — part of how often your brand is recommended in the market as a whole, alongside the six chat engines. Same Google AI features, different job.

Unaided awareness means the model names your brand even though the prompt never mentioned you — for example “best rank tracker for agencies” with no brand in the question. That is the closest AI equivalent of classic brand tracking: whether you are top of mind when a buyer asks the category, not when they already typed your name.

It is a single 0–100 figure from four published parts: unaided mention rate at 40%, average recommendation position at 35%, citation rate at 15%, and surface coverage at 10%. The weights are shown beside the total, so a drop always points at a specific part of the formula rather than a black box. Category rank (#4 in “SEO software”) is the ordinal on the leaderboard; the score is this composite.

When an AI answer lists several brands, position is where you sit in that shortlist — first, second, third. A brand mentioned first 40% of the time outperforms a brand mentioned last 70% of the time, because position drives attention. The index weights frequency by that position instead of treating every mention as equal.

Co-occurrence counts how often two brands appear in the same AI answer. It shows which competitors the model treats as your real set — the brands named beside you — and which ones appear in answers where you are missing. That pairing list is more useful than a static competitor spreadsheet, because it comes from what the models actually say.

Cited sources are the domains AI platforms use as references when they answer questions in your industry. A high citation share means models lean on that site to form answers. The index lists the top sources for your category so you can see who earns authority in AI search — and whether your own domain is among them.

Weekly. Category prompts are re-run once a week across the tracked surfaces. Model answers drift over days rather than hours; a weekly series gives a trend you can act on instead of noise you have to smooth. History starts when you add the category and accumulates while the project is active.

Yes. Generate the index for a wide market (“SEO tools”) and for a specific segment (“rank tracker for agencies”, “local SEO software”). Preferences change by audience and price point; measuring both stops a strong niche rank from being averaged away inside a broad leaderboard.

The index is included on every paid plan, from the $39 Starter upward — there is no separate AI Visibility Index subscription and no enterprise-only gate on the module itself. What the plan decides is the wider project limits (prompts, competitors, reports) that sit beside it.

Semrush publishes an enterprise AI Visibility Index study (large prompt corpora, industry leaderboards, source analysis) and sells AIO as Semrush for Enterprise — pricing is demo-based, not a public list price. Evertune’s AI Brand Index is a 0–100 unaided brand score with competitive benchmarking on a dedicated GEO platform, also sold via demo. TrueRanker puts a category index — unaided rank, average position, sources and co-occurrence — inside the same project as your Google rankings, from $39/mo, without a second invoice.

Raise unaided mentions and citation share in the category: publish answer-shaped content that models can quote, earn mentions from the sources your niche already cites, and close gaps where co-occurrence shows a rival in answers you miss. The index measures the outcome; the tracker’s stored answers show which prompts and arguments to fix first.

ChatGPT, Gemini, Claude, Perplexity, Grok and DeepSeek, plus Google AI Overviews and AI Mode as category surfaces. That set matches how buyers actually ask — chat assistants and Google’s AI answers — without claiming a proprietary multi-hundred-million prompt corpus.