AI search visibility: are you mentioned or cited when buyers ask?
A sampled check of whether AI answers mention or cite your site.
- OpenRouter keyKeyYour own key. Managed AI credits are not offered.
- CostCostModel tokens plus a per-request web-search fee, charged by OpenRouter, for each question.
- ModelNo keyA web-search capable model. The current setting uses Gemini 3.5 Flash with the web plugin.
- 5
- questions at most
- 3
- signals per question
- 1
- click to run
- 1
- model per check
AI search visibility · last check
5 questions| Buyer question | Mentioned | Cited | Rivals named |
|---|---|---|---|
| Which app reminds me to water my plants? | Yes | Yes | Leafnote, Greenlog |
| What is the best plant care app for beginners? | No | No | Leafnote |
| How do I track watering for many houseplants? | Yes | No | Greenlog |
| Is there a free plant reminder app? | No | No | Leafnote, Greenlog |
| Alternatives to Leafnote for plant care? | Yes | Yes | Greenlog |
5 model runs and 5 web-search requests on your OpenRouter account. Runs on click only.
One model at one moment. Treat it as a sample, not a ranking.
How the check runs
Choose the questions
Write up to five buyer questions or accept the suggested ones from your profile.
Confirm the cost
Each question uses model tokens plus a web-search fee on your OpenRouter account.
Read the answer table
For each question see whether you were mentioned, whether your site was cited, and which rivals appeared.
Act on gaps
A missing mention is a prompt to look at the pages an answer cited and what they cover that yours does not.
| OpenRouter key | KeyYour own key. Managed AI credits are not offered. |
|---|---|
| Cost | CostModel tokens plus a per-request web-search fee, charged by OpenRouter, for each question. |
| Model | No keyA web-search capable model. The current setting uses Gemini 3.5 Flash with the web plugin. |
- Promise to get you cited. It measures; it does not influence an assistant.
- Present a score as a rank. There is no universal position in an AI answer.
- Run on a schedule or spend credits without a click.
BlogFactory never publishes live. Agents propose, people approve, and the furthest an agent goes is a CMS draft.
More: what you get, what changes, limits and questions
People increasingly ask an assistant what to buy or use. This check asks a web-search model up to five questions a buyer would ask and records, for each, whether your brand was mentioned, your site cited, or rivals named. It is a sample from one model at one moment, so it reads as a spot check, not a ranking.
What you get
Buyer questions, not keywords
Up to five questions you would expect a buyer to type, for example the best app for a task.
Three signals per question
Mentioned, cited and rivals named, recorded from the answer and its sources.
Stored per site
The last result is stored on the site and replaced by the next check, so you can reopen it without paying again.
What changes
Before
- You wonder whether assistants mention you and cannot tell.
- Asking by hand gives one answer you cannot compare.
- Rivals named in answers go unnoticed.
With BlogFactory
- Up to five buyer questions asked of a web-search model.
- Mentioned, cited and rivals named recorded per question.
- The last result stored on the site; costs shown before you click.
OutcomeA sampled spot check, not a ranking.
Limits to know
- One model, one moment. Answers vary between runs and between assistants.
- Only the last result is stored; this is not a history or a trend line.
- It does not test every assistant. A mention in one tells you little about another.
Questions
Can BlogFactory make AI assistants cite my site?
No. It measures whether one web-search model mentions or cites you for a few questions. What influences citations is outside its control.
What does a check cost?
Model tokens plus a web-search fee per question, charged to your OpenRouter account. The panel states this before it runs, and exact prices are on OpenRouter's pages.
Is the result repeatable?
Not exactly. Language models vary between runs, so treat a single check as a sample and rerun it before drawing conclusions.
Does it check ChatGPT, Perplexity and others?
It checks one search-enabled model through OpenRouter. It does not query other assistants.