The yes machine
Without live data, a chatbot answers from what it learned in training. It can name competitors that have since closed and guess at demand it never measured.
Vettoz · the idea validation tool inside Claude, ChatGPT and Cursor
Tell your AI the idea: a physical product, a digital product, a SaaS or an app. Vettoz reads what buyers say on Reddit and X, what gets watched on TikTok and YouTube, which ads have run for months and what competitors charge. He scores the idea and shows the evidence behind each score, and your AI writes the report: price, keywords, names with free domains and a test you can run this week.
01The problem
Ask a chatbot if your idea is good and it usually finds a way to say yes. It hasn't read this year's threads, seen which ads keep running or checked what the competition charges. So the idea gets built on a hunch, and the market answers months later.
Without live data, a chatbot answers from what it learned in training. It can name competitors that have since closed and guess at demand it never measured.
Doing it by hand means an afternoon or more across Reddit, TikTok, the Ad Library, Amazon and the App Store, with tabs of notes and no score at the end.
You pick a name, design the logo, and then find every domain taken. It happened to us: this product was called Rekon until a check found its .com, .io, .ai, .si, .app and .dev all registered.
02How it works
Vettoz connects through MCP, the standard Claude, ChatGPT and Cursor already support. He runs no writing model: he collects the evidence and scores it, your AI turns it into the report, in your language.
"Check this idea with Vettoz." Say what it is, who buys it, and whether it's a physical product, a digital product, a SaaS or a mobile app.
Reddit threads and their comments, X posts, TikTok and YouTube videos, active Meta ads, Google results, and TikTok Shop or the App Store depending on the kind.
Pain, demand, willingness to pay, room and reach, each scored by Jev, TypeSafe's judgment model, with the evidence linked. Flags when big players own the market or the evidence is thin.
The verdict with quotes and links, the competitors and their prices, keywords, names checked against free domains, branding, and a test you can run in 7 days. Saved in your account to read and share.
A score is a reason to look closer, not a promise. Every item Vettoz returns carries its link, and the report method tells your AI to cite them and to say so when a source found little, instead of filling the gap.
03Works with
No new app and no second AI subscription. Your AI, on your plan, writes the report. Vettoz brings the research and the score.
Claude.ai, the desktop app, the mobile app and Claude Code. Add a custom connector and sign in.
A custom connector in ChatGPT's settings. Sign in with your Vettoz account.
One block in your MCP settings file, with a key from your account.
Other AIs that support remote MCP servers connect the same way.
04Pricing
Plans count checks because every check pays for live data from Reddit, TikTok, YouTube and the Ad Library. Free checks never expire; paid checks reset on the 1st of each month, and a check that finds nothing is given back.
Fifteen checks a month, for founders testing what to build next. Yearly saves $38.
Deeper checks and more of them, for people who test ideas every week. Yearly saves $78.
05FAQ
MCP (Model Context Protocol) is how AI assistants use outside tools. Once Vettoz is connected, your assistant calls him by itself when you ask about an idea, the same way it searches the web or reads a file. You never leave the chat.
Any AI that supports MCP connectors: Claude (web, desktop, mobile and Claude Code), ChatGPT, Cursor, VS Code and more. Your account has setup steps for each one.
Four kinds: physical products (he adds TikTok Shop products and Amazon results), digital products sold once like templates and courses (Etsy, Gumroad and similar), SaaS (software sites and alternatives pages) and mobile apps (App Store apps with their price, rating and rating count).
Reddit threads and their top comments, X posts, TikTok and YouTube videos, active ads in the Meta Ad Library, Google results, TikTok Shop and the App Store. Every item comes back with its link, so you can open the source.
Vettoz scores five things from the evidence with Jev, TypeSafe's judgment model: how painful the problem is, how much demand shows, whether people pay, how much room a newcomer has, and how easy the buyers are to reach. Demand also counts views, comments and ads in code. Big entrenched players cut the score, and thin or off-topic evidence keeps the verdict at test at most. Every dimension comes with the reason for its score.
Your AI does, from the evidence and a fixed method Vettoz hands over: verdict, quotes, competitors, price, keywords, names, branding, next test and risks. Vettoz runs no writing model. The report is saved in your account.
Vettoz asks the registries directly (RDAP, or WHOIS for .io, .co and .me) and only says available when the registry has no record. That means unregistered at that moment, not reserved for you: register the one you like soon.
2 free checks, once. Starter has 15 a month and Pro 50 a month, and reads 6 keywords and 6 Reddit threads per check instead of 3. A check that finds nothing doesn't count.
No score can promise sales. Vettoz tells you what the market says right now, with links, so you can decide and run a cheap test first. The report method ends with that test and the number that would mean go.
Yes. Your ideas and reports are visible only in your account. We don't sell them, share them or use them to train AI models. To search, Vettoz sends your keywords to its data providers; to score, the idea and the evidence go to its scoring model. Never your email.
Cancel any time from your account. The plan stays active until the end of the period you paid for, and nothing is charged after that.