OpenAI API
Chat completions via the Responses API when a model and key are configured.
AI visibility
Discover How AI Sees Your Brand
Measure your brand's visibility across AI assistants, uncover competitors, and discover opportunities to improve your presence in AI-generated answers.
Scanning runs from the administrator console. Public visitors cannot queue a scan.
A guest can run a limited scan without an account. Results are stored and shown from the database, never invented.
Example score layout
—
Your score appears after a real scan finishes.
OpenAI
Uses the configured API model
Claude
Same neutral questions
Gemini
Unavailable until a key is set
Anthropic Claude API
Messages API. The model name is an environment or admin setting.
Google Gemini API
generateContent. Unconfigured providers stay empty.
Three steps
1. Crawl
The worker reads a limited number of public pages, honoring robots.txt and refusing private network addresses.
2. Ask
The same buyer questions go to each configured provider. The brand name is not planted in every prompt.
3. Score
The score is the share of successful answers that mention the brand. Failed calls are left out.
Reporting
A finished scan shows the score, provider breakdown, questions, excerpts, competitors found in the answers, and recommendations based on those results and the crawl.
Competitors
Names and other domains that show up in answers are counted. The report highlights questions where those names appear and the scanned brand does not.
What you can do with a report
See mention presence, not a black-box rank
Read which questions included the brand
Compare providers separately
Save domains after you register
Schedule weekly scans on paid plans
Export CSV on plans that include it
Pricing
Prices below are the plan catalog. Billing is not connected, so choosing a plan does not charge a card.
Free
$0
A limited number of scans and the basic visibility report.
Starter
$49
One monitored domain, weekly scans, history, and email notifications.
Pro
$149
Several domains, competitor tracking, custom questions, and recommendations.
Agency
$399
Higher quotas, more domains, and report exports for client work.
FAQ
Is this the same as asking ChatGPT, Claude, or Gemini in their apps?
No. Reports use the provider APIs and the model names configured for this installation. Consumer apps can answer differently.
What happens if an API key is missing?
That provider is marked unavailable. The product does not fill in fake answers, and failed calls are not counted as “brand not mentioned.”
How is the score calculated?
Version 1 is 100 times the number of successful answers that mention the brand, divided by the number of successful answers.