The price locks in. That's the whole deal.
Three prices, one ladder. Every subscriber gets the same 4,000 credits a month; the founders' reward is paying less for it, for as long as they stay subscribed.
Plan first. Ship with real developers.
Swazona is the intelligence-and-execution layer: demand signal, a structured BRD, milestones, and a vetted developer marketplace. Not another prompt-to-app prototype generator.
Founders pricing, locked in for as long as you stay subscribed.
Every tier grants the same 4,000 credits a month. The founders' reward is the price, locked while you stay subscribed, not a bigger grant.
Taxes calculated at checkout.
A deterministic cost table, not a vibe.
Real numbers from the running config, not a marketing estimate. Failed runs are never charged.
| Action | Cost |
|---|---|
| Audit run (platform fee) | 10 credits |
| Tournament run (platform fee) | 15 credits |
| Idea → BRD → stack (full journey) | 80-200 credits |
| Full-project build (from-zero lane) | Credit-metered by project size: no flat number to promise. Cheapest with a connected model account (BYO). |
| Metered AI actions (discovery, BRD, Tech Lead review) | Priced at real model cost × 1.5, ranges by task |
Failed runs are never charged.
Your AI subscription does the thinking; credits meter the machinery.
When you connect your own Claude or Codex account, that account pays for the actual model tokens; Swazona doesn't re-sell them. Credits cover the machinery around it: orchestration, queueing, judged comparisons, and the platform fee on audits and tournaments. When Swazona generates on your behalf with no connected account, that cost is metered to the real model price, plainly, never marked up to look free.
12-15% on each escrow release at milestone-close. On a typical small project ($2,000-$10,000), that's $240-$1,500 per project.