About the build
Built so far
Wisdom Pro is built with AI models doing most of the coding, reviewing and testing. The footer keeps a running count of that work, plus the owner's own time directing it since Oct 9, 2026. It was last updated Nov 21, 2026.
- 1,292 hours
- Two kinds of working time, added together. 1,163 hours are the time AI agents spent building Wisdom Pro, counted from the work logs across 865 work sessions: when nothing happened for more than 15 minutes, that gap is left out, and when several models work at once, each one's time counts. The other 129 hours are the owner's own active time directing that work, measured from the owner's messages to the agents, with gaps over 30 minutes left out. Time spent listening to mixes and testing the app isn't in any log, so it isn't counted: the real total is higher.
- 24.1 billion tokens
- Tokens are the small pieces of text a model reads and writes: roughly three quarters of a word each. The count includes the code and notes the models read (most of it re-read from a cache on every step) and everything they wrote.
- About $8,700 at API prices
- What those tokens would cost at the providers' public per-token API prices (Anthropic for Claude, OpenAI for Codex), rounded. It is not what was spent: the work runs on flat monthly subscriptions, so the real cost is much lower. It is a way to show how much work went in.