Open a tab. Turn off your wifi. Get a story.
The entire AI fits inside the page — nothing leaves your device.
Any browser, any device. No sign-up, no app store, no install. Just a URL.
Eeny (free, instant), Meeny (better stories, unlock once), or Miny (enterprise). All run locally in the tab.
Single story, swarm of many, endless chain, or a full book with a set word count. The AI is already there — no waiting.
Every AI tool you've used sends your prompts somewhere. Servers. Data centers. Third parties. TinyTale doesn't. The model runs on your GPU, in your browser tab. Your words never leave your device — not to us, not to anyone.
"Every AI product you've used is a client-server application. The 'AI' lives somewhere else, on someone else's hardware, paid for by someone else's electricity. We asked: what if the AI lived in the URL?"
That's what TinyTale is. Not a wrapper around GPT. Not an API call. A model that fits inside a webpage — and runs the same way whether you're on wifi, on a plane, or in a bunker with no connection at all.
The technology has been ready. WebGPU landed in Chrome in 2023. Quantized models can now be under 5MB. WASM runs in every browser. What was missing was someone to actually build the thing and put it in front of people.
No account. No server. No bill at the end of the month. Just open the link, pick your mode, and read.