---
title: "Discussion on \"A token is not a word, and your chat box is not the bill\" | Hashnode"
description: "Discussion on \"A token is not a word, and your chat box is not the bill\". When a tool says a request used \"1,000 tokens\", it's tempting to read that as 1,000 words. It isn't. A token is the unit a model's tokenizer cuts text into. Sometimes that's a whole word, sometimes a"
url: "https://hashnode.com/posts/a-token-is-not-a-word-and-your-chat-box-is-not-the-bill/6aec482360b658ec49d14c42"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [A token is not a word, and your chat box is not the bill](https://melonape.hashnode.dev/a-token-is-not-a-word-and-your-chat-box-is-not-the-bill?utm_source=hashnode&utm_medium=feed)

When a tool says a request used "1,000 tokens", it's tempting to read that as 1,000 words. It isn't. A token is the unit a model's tokenizer cuts text into. Sometimes that's a whole word, sometimes a

Nov 5 · 5 min read · melonape.hashnode.dev  
Read the article

-   [#openai](https://hashnode.com/tag/openai)
-   [#ai](https://hashnode.com/tag/ai)
-   [#llm](https://hashnode.com/tag/llm)
-   [#beginners](https://hashnode.com/tag/beginners)

## Written by

[melonape](https://hashnode.com/@melonape)

@melonape

  
[View profile](https://hashnode.com/@melonape)

Full-stack engineer, 10 years building for the web. Writing about AI tools, open source and the small tools I make.

Full-stack engineer, 10 years building for the web. Writing about AI tools, open source and the small tools I make.

## Responses 1

Markdown supported · ⌘↵ to post

[AI Developer](https://hashnode.com/@lycore)

[Nov 5](https://hashnode.com/posts/a-token-is-not-a-word-and-your-chat-box-is-not-the-bill/6aec482360b658ec49d14c42/comment/6aec9d6ae4ee234ef66128ae)

The point that a count from one encoding can't be reused for another model is easy to miss when estimating budgets. One practical addition: for chat-style requests, per-message overhead and tool or function schemas can add a surprising number of input tokens on every turn, so a words-to-tokens estimate on the user text alone tends to under-count. Comparing a local tiktoken estimate against the usage field returned by the API on a few real requests is a cheap way to calibrate that gap. Do you know how much the gap grows on code-heavy prompts?
