Hey all! ☀️
I worked on this project for some hours to scratch my own itch and I think it came up pretty nice. The official OpenAI tokenizer only works up to GPT-3, but I wanted to know the tokens on each model since I've been working a lot with prompts in different apps.
It's a purely front-end application, so the prompts are never stored or transmitted through the internet to anywhere. The analytics are done using Fathom, so it's also privacy-friendly :-)
Hope it's useful to someone else!
Talkzy