Pricing
Prepaid credits — pay per decision, not per month. Credits never expire. No calls, no sales.
Not sure yet? See a real Discovery run first — the real output with its sources, no account needed.
Credit packs
Founder Discovery Pass
5 markets in one go — or one market plus 4 deep-dives on what it finds.
Get creditsMulti-Market
14 markets in one go — or one market plus 13 deep-dives on what it finds.
Get creditsWorking Founder
Every market in our catalog searched in one go — 29 of them, one run each. The working-founder tier.
Get creditsBuy with your risk reversed
- See the price first. Every run shows its exact credit cost before you spend it — no surprise charges.
- A failed run is free. If a run errors out on our side, its credit is refunded automatically.
- Credits never expire. Buy what you need and use it whenever. No subscription, no sales calls.
- 14-day pack refund. Unused credit packs are refundable within 14 days.
- It's willing to say no. A weak market hands back nothing, with the reasons — you keep the rationale, not a build you'd regret.
What one credit gets you.
One credit buys one answer: Build it, Test it first, or Drop it — with the reasons and one thing to do next. A Discovery run spends a few credits to go and find ideas, each backed by a problem you can click through to. A weak market comes back with nothing, and why.
How credits work: most runs cost 1 credit; Compare is 1 credit per 3 ideas (rounded up); a Discovery run is 2 credits for the first market, +1 for each extra one. New accounts get 2 free credits for any run except Discovery — a Discovery run always comes out of a paid pack.
What one run actually does
It plans your market into sub-markets, searches live — search engines, listing sites, public APIs — reads the pages it is allowed to read, and puts up to 150 candidates through fixed deal-breaker checks before handing back at most 5. The run we publish made 60 searches, read 24 pages and turned down 36 of 69 candidates, each with a named reason.
See exactly what it does, step by step →Why not just ask a chatbot?
- It is built to say no. A chatbot tends to agree with you; here the default answer is drop it.
- The score is arithmetic. The model rates each area; the 0–100 total is computed in code from a fixed checklist, so it cannot talk itself into a pass.
- Sources you can open. On the run we publish, 32 of 56 problems link to the page we found them on, and the rest are marked not verified.
- A record that grows. Every Discovery run adds to one shared library of recorded problems that later runs draw on.
Still comparing?
Read a real idea report, see what one idea check looks at, or browse the pain library. Weighing us against something else? See WhittleOS vs ChatGPT, WhittleOS vs IdeaProof, WhittleOS vs FounderPal and our own gate run against eight ideas with known outcomes. Can your idea carry a price? The free customers-needed calculator does that arithmetic — no signup.
What's NOT included: No calls. No coaching. No done-for-you. Pure product. Still have a question? Read the FAQ.
Payments are processed by Polar, our Merchant of Record. Prices in USD. Unused credit packs are refundable within 14 days; a failed run auto-refunds its credit — see Refunds.