Market research and competitive intelligence for founders, and the white-label engine agencies, accelerators and consultants use to pre-qualify inbound startups.
What is Preuve AI?
Preuve AI is a web-based startup validation tool. You describe your startup idea in one or two sentences; it runs 10 AI agents against 60+ live sources — Crunchbase, Google Trends, Reddit, Product Hunt, G2, and others — and returns a viability score, up to 15 profiled competitors with funding and pricing, and a list of blind spots. The product runs in the browser, is built by an independent solo founder, and is powered by Claude Opus 5. As of the page's data, over 7,000 startup ideas have been analyzed. It is backed by Google for Startups, AWS, Cloudflare, and Exa.
Key Features
- Viability Score — Unbiased 0-100 score cross-checked by multiple AI runs; in a July 2026 benchmark of 6,000+ ideas, only 17.5% scored 70 or above, so a high score is earned.
- Competitor Mapping — Up to 15 competitors per report with live pricing, funding rounds, user counts, and weaknesses from sources like Crunchbase and public pricing pages.
- Source-Linked Claims — Every number in a report links back to its source (Reddit, X, TechCrunch, SEC EDGAR, etc.), so nothing is invented.
- Pivot Recommendations — The report suggests 3 pivots to improve your score, with an average lift of +10 points.
- Market Sizing — Estimates TAM/SAM/SOM using data from Statista, industry news, and market intelligence databases.
- Radar Pro — A subscription tier at $19/month that re-scans your report monthly, flags new threats, and tracks score changes.
- MCP Server & Agent API — Lets you run a full validation from Claude Code, Cursor, or any MCP client.
Who is it for?
- First-time founders — use the free 60-second scan to decide whether an idea is worth building before investing six months.
- Serial founders — test multiple ideas and compare them side by side with competitor intelligence.
- Solo builders and vibe coders — get a harsh, data-backed reality check on new concepts instead of a flattering answer from a generic LLM.
- Founders preparing for fundraising — the Investor-Ready Package generates an executive summary, investment memo, and audio pitch for investor meetings.
What can you do with Preuve AI?
- Pre-screen ideas: Run free scans on any idea to get a go/no-go verdict, main risks, and 2 competitor previews in about 60 seconds.
- Iterate and re-run: After buying a report, tweak your positioning, re-run the analysis, or correct a section (like a persona or channel) and it recalculates without extra credits.
- Monitor your market: Subscribe to Radar Pro to re-scan your report monthly and get alerts when competitors raise funding or launch new features.
- Generate investor materials: Produce a business plan, investor-ready documents, and an audio pitch directly from the report.
How does Preuve AI work?
Describe your idea in one or two sentences. Then:
- Ten specialized AI agents run in parallel: market analyst, competitor scout, community researcher, financial analyst, real-time news scanner, and a cross-validator.
- They search 60+ live sources, pull real data on competitors, funding, and trends, and filter out junk.
- The scoring model runs multiple times; if runs disagree, the system reruns until the verdict converges.
- Every claim in the final report links to its source.
Pricing
- Free scan — includes a viability score, main risks, 2 competitor previews, and a go/no-go verdict.
- Founder Report — $29 one-time, with a 14-day satisfaction guarantee; includes full analysis, competitor mapping, pivots, business plan, and source links.
- Investor-Ready Package — $499.
- Radar Pro — $19/month.
- Packs — 5 reports for $95 or 10 reports for $159.
- Lifetime plans start at $499.
FAQ
Is my business idea kept private?
Yes. Your idea is encrypted, processed only by AI, and never shared, sold, or used to train models. No humans review submissions, and you can delete your data anytime.
What do I get in the free scan?
The free scan gives you a viability score, the main risks dragging your idea down, 2 competitor previews, and a clear go/no-go verdict. It takes about 60 seconds.
Which sources are scanned?
Crunchbase, Google Trends, G2, Capterra, Product Hunt, Reddit, Quora, public pricing pages, market intelligence databases, app stores, and more than 30 other live market signals. Every claim links back to its source.
What AI model powers Preuve?
Every paid report runs on Claude Opus 5, Anthropic's most capable model, and the verdict is cross-checked across multiple runs. The free scan also uses Opus 5, per the page.
What is your refund policy?
Eligible report purchases are covered by a 14-day satisfaction guarantee. Depending on the case, the company may offer account credit, a rerun, troubleshooting, or a cash refund. Subscriptions renew until canceled.
