# Markdown is 15% more token efficient than JSON **URL:** \ **Category:** Prompting\ **Tags:** output-markdown\ **Created:** [June 26, 2024, 9:01pm UTC](https://community.openai.com/t/markdown-is-15-more-token-efficient-than-json/841742 "2024-06-26T21:01:34Z")\ **Posts on this page:** 1\ **Showing post:** 4

**Author:** ![\_j](https://sea2.discourse-cdn.com/openai1/user_avatar/community.openai.com/_j/32/766292_2.png) [@\_j](https://community.openai.com/u/_j)\ **Post date:** [June 27, 2024, 3:29am UTC](https://community.openai.com/t/markdown-is-15-more-token-efficient-than-json/841742/4 "2024-06-27T03:29:03Z")

> [@anon22939549](#): > > Markdown is also the “native” language of most LLMs, as such if will tend to tokenize better too. While that may be true for RLHF and supervised tuning, the “native language” on which BPE token encoder is tuned and optimized is learning corpus. Intensely formatted markdown =\> minimized HTML: 309 to 546. (a how-to-use-markdown demonstration document) o200k is less efficient on both the English markdown and the HTML. > **Here is an example document in Markdown, then HTML** > > ### Markdown Example Document > > ````auto > # My Markdown Document > > ## Introduction > > Welcome to this example document. It demonstrates various Markdown formatting types. > > ### Formatting > > **Bold text** can be created using double asterisks or double underscores. > > *Italic text* can be created using single asterisks or single underscores. > > ~~Strikethrough text~~ is also possible using double tildes. > > ### Lists > > #### Unordered List > > - Item 1 > - Item 2 > - Subitem 2.1 > - Subitem 2.2 > - Item 3 > > #### Ordered List > > 1. First item > 2. Second item > 1. Subitem 2.1 > 2. Subitem 2.2 > 3. Third item > > ### Links and Images > > [OpenAI](https://www.openai.com) > > ![OpenAI Logo](https://www.openai.com/assets/images/openai-logo.svg) > > ### Code > > Inline code: `print("Hello, World!")` > > Block of code: > > \```python > def greet(): > print("Hello, World!") > \``` > > ### Blockquotes > > > This is a blockquote. It can span multiple lines. > > ### Tables > > | Header 1 | Header 2 | > | -------- | -------- | > | Row 1 | Data 1 | > | Row 2 | Data 2 | > > ### Horizontal Rule > > --- > > ### Task List > > - [x] Task 1 > - [] Task 2 > - [] Task 3 > > ```` > > And here is the HTML version of the same document: > > ### HTML Version > > ```auto > > > > > > My Markdown Document > > > >

My Markdown Document

> >

Introduction

> >

Welcome to this example document. It demonstrates various Markdown formatting types.

> >

Formatting

> >

Bold text can be created using double asterisks or double underscores.

> >

Italic text can be created using single asterisks or single underscores.

> >

~~Strikethrough text~~ is also possible using double tildes.

> >

Lists

> >

Unordered List

Item 1
Item 2 >
- Subitem 2.1
- Subitem 2.2
>
Item 3

> >

Ordered List

First item
Second item >
1. Subitem 2.1
2. Subitem 2.2
>
Third item

> >

Links and Images

> >

OpenAI

> >

Code

> >

Inline code: print("Hello, World!")

> >

Block of code:

> >

def greet():
> print("Hello, World!")
>

> >

Blockquotes

> >

>
This is a blockquote. It can span multiple lines.
>

> >

Tables

> > > > > > > > > > > > > > > > > > >

Header 1	Header 2
Row 1	Data 1
Row 2	Data 2

> >

Horizontal Rule

> >

Task List

> >

[x] Task 1
[] Task 2
[] Task 3

> > > > > ``` > > ## Minimized HTML > > ```html >