Playground
A tiny writing lab for prompts
This page is intentionally small. It is not a full-featured tokenizer. It is just enough UI to notice that when your writing gets more polite, more structured, or more verbose, the token count quietly goes up.
What the model feels
Rough estimate only; real APIs use provider-specific tokenizers.
Length
~37 tokens
Every polite sentence, bullet list, or extra example lives here.
Shape
6 lines
Models do not see paragraphs, but structure still nudges behavior.
Tradeoff
cost 22 tokens
Try rewriting the same ask in fewer words and notice how the estimate shifts.
If you want real numbers, use your provider's tokenizer API. This page is here to build a small habit: when you tweak a prompt, glance at the length and ask yourself whether the extra words are buying you better behavior or just a higher bill.