Token Chunker

Paste long text, pick a model, get clean token-aware splits with optional overlap. The chunks a language model actually wants. Everything runs in your browser. Nothing is sent anywhere.