See how language models split your text into tokens — count tokens, inspect token IDs, and estimate API costs
Prices are ballpark figures for intuition — check each provider's current pricing page before budgeting.
Large language models don't read characters or words — they read tokens, subword units produced by byte-pair encoding (BPE). Token count determines API cost, context-window usage, and why models struggle with character-level tasks. This playground uses a GPT-4-style cl100k tokenizer running fully in your browser: hover any chip to see its token ID, and notice how spaces attach to the following word, emoji explode into several tokens, and rare words split into pieces.
Your data stays private. All processing happens locally in your browser. No data is collected, stored, or sent to any server.