Preuve inside your AI assistant. Connect Preuve to Claude, Cursor, Codex, or any MCP-capable agent. Validate ideas, run scans, pull reports, and generate up to five ideas a day without leaving the chat. Paid tiers unlock deep scans and full exports. Docs at docs.preuve.ai.

Cloneability & Moat. How fast could someone copy this? A new report section gives you a clone-time estimate, which of your defenses hold up, and what happens if a well-funded player moves in.

Pre-mortem on every paid report. Your riskiest assumption stated plainly, the test that would break it, and kill criteria that tell you when to stop.

Verified competitor pricing. Paid reports show what competitors actually charge, quoted from their pricing pages, with the category range and where your price would land.

Assumptions disclosed. Reports now show what the scan inferred about your business, not what you stated, so you can correct a wrong assumption instead of arguing with the score.

Hide a competitor. Wrong match, not a real rival, not relevant. Hide any card with a reason, and it stays hidden in exports and client shares.

Idea Studio. The idea composer is now a dedicated page with tier badges, saved toggles, and generated ideas that survive navigation.

Go-to-market channel plan. Founder Fit now recommends which acquisition channels suit you, and flags when its advice is based on stale profile answers.

More countries. Paid reports can target up to three extra countries alongside the main one. Coverage now includes Bangladesh, Pakistan, Sri Lanka, Nepal, Ghana, Ethiopia, Morocco, Algeria, and a Global option.

Progress you can walk away from. A running scan shows its progress from anywhere in the app, so you can leave the page and come back.

Subprocessors page. A public list of every vendor that touches customer data, linked from the privacy policy.

Claude Opus 5. Every paid scan runs on the strongest model.
Sharper scoring, tighter reasoning, verdicts that hold across runs.

A faster app. The app opens faster, reports load faster, and status updates arrive sooner. Consultant reports now deliver in about five minutes instead of eight.

Market figures you can check. A figure is only labeled "verified" when the source we cite actually contains it. When the market math collapses to a number that cannot be right, the report recalculates it instead of printing it.

Sharper competitor cards. The report no longer repeats a competitor's own marketing claims as fact, and checks metrics before any number reaches the card.

Financial section redesign. A rebuilt Raise Plan card, a burn breakdown you can read at a glance, and clearer revenue and break-even subsections.

Refine without losing your work. Edit your idea straight from the report menu, keep drafts between attempts and retry a refine that failed.

Stronger demand signals. Proof of Demand surfaces more prospects, lets you rescan whenever you want, and says plainly when a niche has little public signal to find.

Lender-ready Business Model Canvas. The BMC export now targets the audience that reads it: banks, grant committees, and loan officers.

Tighter rate limiting on sign-in, checkout, report, and payout routes.

A blocked signup now explains why it was blocked and gives you a way to recover the account.

A brief network hiccup mid-scan no longer throws you to the crash screen.

Verification-code errors no longer all read as "expired code".

Account deletion now completes reliably, including on retry.

Reports stuck mid-scan are now detected and restarted automatically.

Founder Fit keeps late-arriving answers and warns when a saved plan is based on outdated inputs.