Research & Reading

Preuve AI

The scanner that checks 50+ real sources to tell you if your business idea holds up.

Free scanLevel beginner7 min read
* Independent, free analysis. This content is educational and is not sponsored or paid for in any way. We show the real strengths and the real limitations — an honest review, so you can decide.
What it is

An X-ray of the market before you spend a dollar building

Preuve AI tests whether a business idea is worth building before you go build it. You describe the idea in a text field, and behind the scenes ten AI agents go dig up real data across 50+ places — Reddit, Crunchbase, Google Trends, G2 — to hand you a 0-to-100 score, the competitors that already exist, and the risks nobody told you about. The difference from a regular chatbot: here every claim comes with a link to where it came from.

Why it matters

42% of startups die for the same reason: nobody wanted it

You (or your client) already spent a whole weekend polishing the pitch for a new idea. Everyone who heard it thought it was "amazing." And that's exactly the problem: friends and family almost never tell you the truth.

42%

That's the share of startups that die for exactly that reason, according to CB Insights — not lack of money, not bad execution, but because the market simply didn't want it. Preuve AI targets that exact blind spot: it cross-checks competitor data, search demand and real complaints before you commit time and money.

In your day

Three moments where it's worth running the scan first

Consultant

A client shows up with an idea and asks for your professional opinion. Instead of promising a market report in two weeks, you run the scan on the spot and leave the meeting with a score, real competitors and a first draft of a plan.

Small business owner

Thinking about launching a second product line, but don't want to repeat the mistake of investing without knowing if anyone will pay for it. The report shows whether the "gap" you spotted is real or just a feeling.

Manager

Needs to defend a new idea in front of leadership. Instead of a pretty slide deck full of opinion, you bring a scorecard with real competitors and sources attached.

Step by step

From a rough idea to a source-backed verdict

// interface checked in July 2026 — if something looks different, look for the same goal inside preuve.ai/app
  1. Go to preuve.ai and click "Get my verdict"

    That takes you straight to preuve.ai/app, no account required upfront.

  2. Describe your idea in the "Describe your idea" field

    Two or three sentences are enough, but the field takes up to 50,000 characters — and Preuve itself says more detail means sharper insights.

    the more specific the audience and the pain, the better the result
  3. Fill in "Your situation" (optional, but worth it)

    Your experience (first time or done it before), what stage you're at, launch budget, ideal first customer and target region.

  4. Click "Get My Verdict"

    In about 60 seconds you get the viability score, 2 sample competitors and the go/no-go verdict.

    you only get 2 of these free runs per account
  5. If it's worth going deeper, buy the Founder Report ($29)

    Unlocks all 15 competitors, the go-to-market plan, the 3 pivots and the full business plan, with a 14-day guarantee.

Copy and use now

A description ready to paste into Preuve's field

idea description Idea: [product name or one-liner]. Solves [the specific pain point] for [a well-defined target audience, e.g. accounting offices with up to 5 employees]. Today that audience solves it with [current alternative — spreadsheet, WhatsApp, competitor X], losing [estimated time or money per week]. The revenue model would be [monthly subscription / commission / one-time sale], around [approximate price]. I've already talked to [number] people in that audience and heard: "[a real quote about the pain point]".

Paste this with the brackets filled in into the "Describe your idea" box — it's roughly the level of detail Preuve itself asks for to sharpen the score.

What few people know

What pushes the score up — and what trips it

Do this

  • Write the real pain point and the specific audience, not just the product name — a bigger description means a sharper analysis.
  • Fill in the optional fields (budget, experience, region) — they change the result, because the score also weighs your founder profile.
  • Spend one of your 2 free scans testing a competing idea of yours before deciding which one is worth paying to go deeper on.

Avoid

  • Expecting the report in Portuguese — both the interface and the returned text are 100% in English.
  • Counting on the free scan forever — it's only 2 total per account, after that it's $29 to keep going.
  • Treating "no competitors found" as absolute truth for a very local niche — most sources (Crunchbase, G2, Product Hunt) are heavily US-centric.
Challenge · 5 minutes

Run the idea that's been collecting dust in your notebook

Grab that business idea you (or a client) have been putting off for months. Describe it in 2 or 3 sentences in Preuve AI's free scan and read the viability score.

You've succeeded if, in under 2 minutes, you get a 0-to-100 score and at least 1 real competitor listed — with a link to check it.

Before you spend the $29

My honest score after running two real ideas

3.8

score 3.8 out of 5

I ran two ideas I already had shelved: one scored 61, the other stalled at 38 — and that one stung a bit, because I liked it more. But the report surfaced 2 competitors I didn't even know existed, complete with pricing and funding, and that alone would have been worth the check. Preuve AI's biggest strength is that it doesn't make things up: every number links back to a source, something a regular chatbot alone won't give you. It just gets expensive if you only want to "test an idea for fun": paying $29 per scan works fine once, but validating many ideas adds up fast. And if your idea targets a very local, Brazil-specific niche, be skeptical of "no competitors found" — it might just be because its sources are American.

Tested and checked against official sources on July 16, 2026.

What works in its favor

  • Every competitor, funding figure and demand signal links to a source (Reddit, Crunchbase, G2, Product Hunt) — you can click and verify it, it's not just the model "guessing"
  • The free scan delivers a viability score and 2 sample competitors in about 60 seconds, no card required
  • The report runs on Claude Opus 4.8 with multiple runs cross-checked until the verdict converges, and it's far cheaper than a market research consultant — which the site itself estimates at $5K to $25K for the same deliverable
  • The paid report has a 14-day satisfaction guarantee, and there's even a lifetime plan (Lifetime Pro, $499 one-time) for anyone validating ideas nonstop

Where it struggles

  • Zero Portuguese: interface, form and report all come out 100% in English, with no language selector at all
  • The free tier is short: only 2 scans per account, and even then you only see 2 sample competitors — the rest, up to 13 more, stay locked behind the $29 report
  • Sources are heavily American (Crunchbase, G2, Product Hunt, SEC, Y Combinator) — for a hyperlocal Brazilian idea, the "market gap" it flags may not reflect reality here
  • Pricing only in US dollars, no local plan, and the price ladder climbs fast: the investor package costs a separate $499
How much it costs

From the free scan to the founder report, in dollars

Is it free? Only to start. The 60-second scan doesn't ask for a card, but you only get 2 runs per account — to go deeper on an idea, the full report starts at $29.

Free

$0 — 2 scans total: viability score, 2 sample competitors and go/no-go verdict. No card.

Founder Report

$29 one-time — 15 sections, up to 15 competitors, go-to-market plan, 3 pivots, business plan and linked sources. 14-day guarantee.

Packs

$95 (5 reports) or $159 (10) — works out to $19 or $15.90 each, for testing several ideas.

Radar Pro

$19/mo — re-scans every 30 days for founders actively building. Cancel anytime.

Investor-Ready

$499 one-time — pitch deck, memo and financial model, hand-built by the founder, for raising investment.

Lifetime Pro / Business

$499 or $999 one-time — 10 or 20 deep scans per month forever, limited spots.

Important detail: official prices from the pricing page as of July 2026, always in US dollars — the site has no local-currency billing option.

Who it's for

Who the viability score actually serves

Worth it if you…

  • Are a consultant or run a small agency and need to hand clients a quick, sourced market snapshot — there's even a white-label plan built for this (Preuve for Teams)
  • Already have 2 or 3 business ideas in mind and want a quick, evidence-based filter before betting your time on one
  • Are going after investors and need a pitch deck and memo with numbers that hold up in the room

Think twice if you…

  • Just want to "play around" validating ideas for free — the free tier runs out fast (2 scans) and the rest is paid in dollars
  • Aren't comfortable reading in English — there's no translation at all, not even partial
  • Your idea is hyperlocal (a neighborhood service, say) — its sources are better suited to the US SaaS market than corner-shop retail
If this isn't quite it

Other doors for anyone who just wants a quick validation

ChatGPT

To draft the idea, get a first opinion or simulate a "customer" conversation — free, but with no real sourcing behind it

See the lesson →

Perplexity

When the question is about a market trend or recent news and you'd rather search the open web yourself

See the lesson →

Consensus

If validation depends on peer-reviewed academic research — a health or education product, say — it searches papers, not Reddit

See the lesson →

NotebookLM

If you already have interviews and spreadsheets from your own customers and just want to organize that material, it reads and cites it all — but it won't go hunt for competitors on its own

See the lesson →

Common questions

What everyone asks before describing their idea

Is Preuve AI free?

There's a free part: the 60-second scan, with a viability score, 2 sample competitors and a go/no-go verdict, no card required. But you only get 2 runs per account — to see the full report, with all 15 competitors and the business plan, the Founder Report costs $29 one-time.

Do I need a credit card?

Not for the free scan — the form only asks for your idea description (and a few optional fields) before showing the score. A card is only needed when you buy the Founder Report or any paid plan, at Stripe checkout.

Does it work in Portuguese?

No. The interface, the form and the report it returns are 100% in English, and there's no language selector on the site. You can still describe your idea in Portuguese, but the result comes back in English.

Does it work on mobile?

Yes, through your phone's browser — the site is responsive. There's no native Preuve AI app on the App Store or Google Play.

Is my idea kept private, or does someone read what I write?

According to Preuve AI's own security page, your idea is encrypted, processed only by AI — no human reads what you write — and never used to train a model. You can delete your data anytime.

Can I get a refund?

Yes. The Founder Report is covered by a 14-day satisfaction guarantee; depending on the case, support offers account credit, a rerun, or a cash refund.

Do I need to already have a registered company to use it?

No. It's built for people who still just have an idea on paper — one of the form's own fields literally asks "where are you now", with the option "just an idea".

Every day, a new tool explained.

Get it on WhatsApp