Tech in Five - Tokens & Tokenomics

John Savill gives a quick, practical overview of tokens and tokenization in LLMs, how models generate tokens, why context windows matter, and how token usage affects both user experience and cost.

Overview

The video explains the basics of how large language models (LLMs) work with tokens, including how text is converted into tokens (tokenization), how token usage can vary, and how models generate output token-by-token.

It also covers why the context window is a key constraint (how much prior text the model can consider), how reasoning can influence token usage, and why tokens are directly tied to cost in usage-based AI pricing models.

Chapters

• Introduction
• What AI models need
• Turning text into tokens
• Tokenization
• Token use may vary based on many factors
• How models generate tokens
• Context window
• Reasoning
• Token cost
• Summary
• Close