A status for each engine
One line per engine says where you stand: it recommends you, mentions you, names others instead, or names no one. Each square is one of your questions.
When people ask AI, does it name you?
Keep a short list of the questions your buyers ask. LopeBase asks ChatGPT, Claude, and Gemini each one, reads every answer, and shows whether you are mentioned, recommended, and described correctly, and who gets named instead.
On the AI visibility tab of the SEO page. Every check shows its cost in AI credits before it runs.
AI visibility
Three engines. Your questions. Every answer read.
A status for each engine, a trend over time, and the full text of every answer, so you can check the reading yourself.
Every answer, in full
Open any answer to read the exact words. LopeBase marks whether you are mentioned and recommended, where you rank in the answer, whether it describes your product correctly, which competitors it names, and which sites it points to.
Questions you choose
Up to 10 per property, the way a buyer would ask. LopeBase can suggest them from your site, and you can edit every one.
A trend, check by check
Each check is kept, so you can see whether a new page or a launch moved how often each engine names you.
Write the page that gets you named
A question where you are missing can become an article topic for the content autopilot, so you write the page that answers it.
On demand or on a schedule
Run a check when you press the button, or let it run on its own every week or every month.
The cost, up front
Each check shows what it will spend in AI credits before it runs. How AI credits work.
Readable by your agent
Claude, Cursor, or any MCP client can read the results through LopeBase's MCP server. Read the AI visibility guide.
Competitors in AI answers
See who gets named when you do not
Every brand the engines name for your questions, gathered from the same answers. No extra check, and no extra cost.
- Every brand namedEach competitor the engines mention for your questions, found from the answers themselves, not from a list you have to keep.
- How often, and how highHow many answers name each one, and its average position in them, next to your own.
- The reasons givenWhat the answers say about each brand, like a free plan or a feature you lack, so you know what to answer on your own pages.
- Tracked over timeYour share beside theirs, check after check, so you can see who is gaining.
How it works
Four steps, and LopeBase does three of them.
- 01
Pick your questions. Up to 10 per property, the way your buyers would ask. Take LopeBase's suggestions from your site, or write your own.
- 02
Ask. Each question goes to ChatGPT, Claude, and Gemini, when you press the button or every week or month on its own.
- 03
Read. Every answer is checked for whether you are named, recommended, ranked, and described correctly, and who is named instead.
- 04
Act. Turn a question where you are missing into an article topic, and see on the next check whether it moved.
What it measures, and what it does not
A reading is only useful if you know how it was taken.
It asks each model directly, without live web search.
LopeBase asks ChatGPT, Claude, and Gemini through their APIs, with no live web search. The ChatGPT, Claude, and Gemini apps can search the web, so what a person sees there may differ. Treat it as a reading of what each model knows about you.
- Costs shown before every checkEach check shows its cost in AI credits before it runs. The competitor view comes from the same answers and costs nothing more.
- Always LopeBase creditsOne AI key cannot reach all three engines, so checks use LopeBase's AI credits, even in a workspace with its own key.
- Your questions, not all of themIt asks the list you keep. It does not guess at every question people might type.
- The answer, not just a scoreEvery status links to the full answer, so you can check the reading before you act on it.
Discussions, in beta
Buyers also ask people. Hear what hurts.
Discussions finds forum threads about the problem your product solves, on Reddit, X, Hacker News, and more. Now each thread is scored for fit, and the problems people describe are grouped into pain points.
- A fit score from 0 to 100Every thread LopeBase finds gets a score for how well it fits your product, with the reason. Sort by best fit and start at the top.
- The problems, pulled outEach thread lists the problems the poster describes, in plain words.
- Pain points across threadsA summary groups the same problem from many threads and links each group back to its threads.
- One click to a topicTurn a group into an article topic. Grouping uses AI credits and shows the cost first.
- Nothing posted for youYou open the thread and reply in your own voice. How Discussions works.
Which AI engines does LopeBase ask?
ChatGPT from OpenAI, Claude from Anthropic, and Gemini from Google. Every question on your list goes to all three, and each engine gets its own status.
How often does it check?
When you press the button, or on its own every week or every month. Each check is kept, so you can see the trend over time.
What does a check cost?
It uses AI credits. Each check shows its cost in credits before it runs, so you always know what it will spend. The competitor view is built from the same answers and costs nothing extra.
Why might the ChatGPT app give a different answer?
LopeBase asks each model through its API, without live web search. The ChatGPT, Claude, and Gemini apps can search the web before they answer, so what a person sees there may differ. Treat a check as a reading of what each model knows about you.
Does it work with my own AI key?
Checks always use LopeBase's AI credits, even in a workspace with its own key. One key cannot reach all three engines, so LopeBase asks them with its own access. Your key still runs your other AI features.
How many questions can I track?
Up to 10 per property. LopeBase can suggest them from your site, and you can edit, add, or remove any of them.
What do I do when I am not mentioned?
Write the page that answers the question. A question where you are missing can become an article topic for the content autopilot, and the next check shows whether anything moved.
Can my coding agent read the results?
Yes. Claude, Cursor, or any MCP client connected to LopeBase's MCP server can read them.
What else do you want to get done?
Every goal runs in the same console, on the same plan.
01 / Search
Get found on Google
See the searches you almost win and the pages to fix first, from Search Console and Bing.
Rank trackingSite auditOpportunity map See how it works 02 / ContentPublish without the weekly grind
Plan articles on a calendar. They are drafted in your voice, checked, and published on schedule.
CalendarYour voiceLink checks See how it works 03 / RevenueKnow your MRR and runway
MRR, past due, and cancellations for every Stripe and Polar account, next to cash in the bank.
MRRPast dueRunway See how it works 04 / GrowthTurn visitors into signups
Traffic from every analytics tool side by side, and a review of where people drop off.
TrafficSignupsConversion review See how it works 05 / ReliabilityHear about outages first
Every site checked every 15 minutes, alerts by email or push, and a status note drafted for you.
15 minute checksAlertsIncident review See how it works 06 / AgentsLet your coding agent set it up
Run one command and your agent adds products and connects sources over MCP.
npx lopebaseMCP serverRead-only See how it worksNot sure where to start? Connect one tool and Overview shows what needs you first.
Start freeFind out what AI says about you.
EmDashFree for one product. No card.