Launching today
Nash
Game-theoretic AI token pricing & margin simulator
4 followers
Game-theoretic AI token pricing & margin simulator
4 followers
Most AI startups pick a flat $20/mo tier and get crushed by whale token burn, or switch to pure usage meters and lose users to bill-shock. Nash is a 100% browser-native AI FinOps workbench that simulates 1,000 log-normal agents using Kahneman-Tversky loss aversion and 2-tier self-selection math. Test 18 real LLM routing stacks, parse private CSV logs locally, run a 240-combination Nash grid-search optimizer, and export production TypeScript router middleware.
Free
Launch Team
LoomTalk it through in Loom. Ship with AI
Promoted
Hey Product Hunt! I built Nash (nashps.netlify.app) because watching AI products bleed negative gross margins on flat $20/mo plans is painful and pure usage meters just trigger bill-shock churn. Instead of guessing pricing tiers on a napkin, Nash runs a live 1,000-agent Monte Carlo cohort right in your browser with zero cloud uploads. Test real Frontier-to-Lite model routing (like GPT-4.5 or Claude 3.7 Sonnet paired with GPT-4o-mini), drop in a CSV of your own user logs locally, run the grid-search Nash Optimizer, and grab production TypeScript router code and a VC unit-economics memo. Try selecting GPT-4.5 Preview and running the optimizer. Would love to hear what gross margins you unlock!