Show HN: Waiting for LLMs Suck – Give your user a game

LLM latency is a UX problem nobody talks about enough. Spinner animations are a cope. This project takes the obvious-in-hindsight approach: drop a tiny game in front of users while the model grinds away. Keeps them engaged, lowers perceived wait time, and honestly just makes your product feel less broken. The repo is open source and lightweight enough to drop into almost any web app. No elaborate setup, no SaaS subscription, just a clever distraction layer between your user and the cold reality of GPU queues. The real insight here is that latency is not always a technical problem you have to solve at the infra level. Sometimes you solve it at the attention layer. Simple, useful, and the kind of thing that improves demo conversion rates immediately. -> Best for: indie hackers and SaaS founders shipping LLM-powered features who want a quick UX win without touching their backend.