All Demos
tokenizertokensclient-side
Token Visualizer
See exactly how your text is tokenised before it reaches the model. Each token gets a unique colour — great for understanding prompt costs.
Input Text
0
Tokens
9
Words
44
Characters
Token Breakdown
Note: This uses a heuristic approximation for visualisation. Real BPE tokenisation (as used by LLMs) may split tokens differently — especially for rare words, code, and non-English text.