Free AI-visibility scan

In our own scans, AI engines sometimes answered market-category prompts without mentioning the relevant brand — and some of the pages they cited did not contain the claims attached to them. Our scan checks each engine×prompt answer for two metrics:

• Mention rate — in how many engine×prompt answers your brand actually appears. Our own self-scan (2026-07-06, google_aio + gemini, 11 prompts, 22 answered cells): Daily AI Agents mentioned in 2 of 22. We're behind, and we say so.
• Citation support rate — whether the pages AI cites actually contain the checked claims (word-coverage containment + fetched-page evidence). Our 2026-07-06 check: 6 of 10 cited pages supported the claim.

Sample checked row (from the 2026-07-06 self-scan): engine google_aio · prompt "What is Peec AI?" · claim "Peec AI is a Generative Engine Optimization (GEO) analytics platform…" · cited URL fetched · verdict supported (quote-level containment, coverage 0.88).

The free scan is up to 5 prompts of your choice across the engines we can honestly reach (currently google_aio + gemini; we disclose blocked engines rather than guessing), delivered by email as a simple table of checked rows with fetched evidence — not a dashboard, not a share-of-voice estimate.

AI teams that want our verified citation rows as eval data for their own models or engines: [email protected], subject "eval data".

Request your scan

Data handling: your domain, prompts, and email are used only to run your scan and deliver your report, retained 12 months or deleted on request. Full details: privacy policy.

Pricing

All prices in plain text, on this page — no "contact sales" below the audit tier. Method behind every claim on this page: the published scan report at /geo-report/ (collection conditions, verdict definitions, and row-level examples).

What you are buying is verification with monitoring, not monitoring alone: visibility trackers tell you where your brand appears in AI answers; this pipeline additionally fetches each page those answers cite and records whether it actually supports the claim attached to your brand — a check we have not found offered elsewhere at any price tier (our competitor teardown, official pricing pages, 2026-07-07).

Free scan — $0, one-time

Up to 5 prompts you choose, run on the engines for which we can complete an automated run under the run conditions published at /geo-report/ — currently Google AI Overviews (logged-out search capture) and Gemini (authenticated chat-UI capture). If an engine blocks the run, requires access we don't have, or returns no answer text, it is listed in your delivered table as blocked/no-answer with the observed outcome — never guessed. Each checked row carries: prompt, engine, the claim as captured, cited URL, fetch outcome (HTTP status), a fetched-page excerpt, and a verdict — supported (the fetched page contains the claim), unsupported (it doesn't), or unfetchable (the page couldn't be retrieved as readable content; counted separately, never as false). Free scans are processed first-come, first-served within a bounded daily scan budget; expected delivery is within 5 business days of your request.

Verification — $49/mo (price testing, checkout in TEST MODE)

The same pipeline the free scan and the /geo-report/ sample already run, weekly:

• 10 prompts, scanned weekly on the engines currently supported under the run conditions published at /geo-report/ (2–3 engines at present): one scan window per week, blocked engines retried once then reported as unavailable for that week and excluded from that week's metrics.
• Citation-verification table on claims about your brand: for each checked eligible claim, we fetch the citation used in that row and record whether the fetched page contains the claim (quote-level containment, as defined in the method).
• Hallucinated-citation flags: every checked cited claim about your brand whose fetched source does not support it is flagged in your weekly table (unfetchable is reported separately, never counted as false).
• The row-level checked-citation dataset for your brand — up to the first 3 citation-bearing claims per answer, in answer order (the same deterministic cap the published pipeline uses).

Verification + history and alerts — $99/mo (price testing, checkout in TEST MODE)

Everything in the $49 tier, plus:

• 25 prompts scanned weekly instead of 10.
• Trend history: week-over-week support-rate and mention-rate movement for your prompts, kept as the same row-level records so you can audit any point on the line.
• Email alerts on your brand: a message when a checked cited claim about your brand is unsupported or unfetchable under the published method (a row is also flagged when the fetched page states the opposite of the claim).

Verification audit — $1,500–3,000, one-time

For regulated or enterprise-shaped teams deploying AI systems: a founder-run, NIST-AI-RMF-mapped audit of how AI engines represent your organization and whether the citations attached to those representations hold up — delivered as a written report with the full row-level evidence. Enterprise inbound routes here; we do not sell an enterprise SaaS tier. Email [email protected].

Billing honesty: "returns data" means the engine produced answer text for at least 80 percent of your prompts that week (calendar week, UTC); if fewer than two engines meet that, the week's scan is not billed, and every report keeps the per-engine record you can check this against. Checkout currently runs in Stripe TEST mode — you cannot actually be charged yet; leaving your email starts the conversation and locks the launch price.