Why Token Counts Matter

Part of the Fundamentals section of Coddy's AI Prompts journey — lesson 9 of 23.

Tokens Matter for Three Reasons

Cost. AI services charge per token, with output tokens costing more than input. A chatty prompt generating a long response costs more than a concise one.

Speed. The AI generates output one token at a time. Shorter responses come back faster.

Context limits. Every model has a maximum token capacity per conversation (the context window). Long prompts leave less room for responses.

This is why efficiency matters. "Can you please help me by explaining in great detail what the capital city of France is?" (~20 tokens) gets the same answer as "What is the capital of France?" (~7 tokens) — at lower cost and faster speed.

For reference, GPT-4o costs about $2.50/$10 per million input/output tokens; Claude Sonnet runs $3/$15. A million tokens is roughly 750,000 words, so even casual use adds up.

Try it yourself

This lesson doesn't include a code challenge.