AI & LLMs
Token
The small chunk of text — roughly 3-4 characters — that an LLM actually reads and bills you for.
Reviewed by the RadarTrek editorial team · June 2026
A token is the small text chunk — averaging 3-4 characters in English — that LLMs actually process, not whole words or characters. Common words are usually a single token; longer or rarer words split into several. Every API call is priced and limited by token count, so the same sentence can cost different amounts depending on punctuation, formatting, and language.
Why it matters
- —API pricing and context limits are both measured in tokens, not words — this directly affects cost.
- —Heavily formatted prompts (bullet points, code blocks, tables) use more tokens than plain prose.
- —Non-English text often tokenises less efficiently than English, raising the cost of the same content.
Where to learn this
🎓
Tokens and Context Windows
How LLMs Actually Work course
This is the exact lesson that covers this term in depth — with examples, diagrams, and a hands-on exercise.