📖 Step 9: AI/LLM#250 / 291
LLM Token
LLM Token
📖One-line summary
The smallest unit of text an LLM processes — usually a sub-word piece.
💡Easy explanation
The 'text chunk' unit an AI processes. Usually shorter than a word; a Korean character takes about 1–2 tokens.
✨Example
How a sentence breaks into tokens
The quick brown fox 점프한다
Total: 7 tokens
⚡Vibe coding prompt examples
>_
Write a Node.js script that estimates token counts for a 10,000-character Korean text on GPT-4o, Claude, and Gemini.
>_
Give 10 practical tips to reduce LLM API cost — system prompt compression, output length caps, and similar.
>_
Design a multi-tenant SaaS backend schema for tracking and billing per-user token usage.
Try these prompts in your AI coding assistant!