JSON to TOON Converter

Shrink structured data for LLM prompts, compare token usage, and copy TOON instantly.

Save 30–60% tokens*

Format style:

Indent width: spaces

All processing happens in your browser

1. Paste JSON

2. TOON output

What is TOON Format?

TOON (Table-Oriented Object Notation) is a compact data format designed specifically for LLM/AI prompts.

Unlike verbose JSON with quotes/brackets/braces (30-60% of characters), TOON uses:

id|name|country|plan
1|Alice|IN|pro
2|Bob|US|free
3|Carol|UK|business

vs JSON (3x longer):

[{"id":1,"name":"Alice","country":"IN","plan":"pro"},{"id":2,"name":"Bob","country":"US","plan":"free"}]

Why Save LLM Tokens?

1 token ≈ 4 characters ≈ $0.0001-0.01 (depending on model)

Model	Input Cost/Token	10K Tokens =
GPT-4o	$2.50/1M	$0.025
Claude 3.5	$3.00/1M	$0.03
Llama 3.1	$0.50/1M	$0.005

Real example: 10K JSON → 4K TOON = $0.015 saved per request × 1000 users/month = $15/month

How JSON → TOON Works (Our Algorithm)

Smart table detection + minimal syntax:

1. Detect arrays of objects → Create header row
2. Extract ALL keys across rows → Consistent columns
3. Escape only delimiter chars → No quotes for simple values
4. Nested objects → `key:{nested JSON}` or `key:[]` for arrays
5. Optional indent (0-6 spaces) → Readable when needed

Input JSON:

{
"users": [
{"id":1,"name":"Alice","active":true},
{"id":2,"name":"Bob","active":false}
]
}

TOON Output:

3-Click Workflow

Paste JSON → Auto-validates syntax
Choose delimiter (| , \t) + indent (0-6 spaces)
Convert → See live token savings + copy TOON

Pro tip: Use pipe | delimiter (handles spaces/commas best)

Use Cases (Perfect For)

AI Agent Developers - Shrink database exports 50%+
ChatGPT/Claude Users - Context window optimization
API Payloads - Log storage costs down 60%
Data Analysts - Export CSVs without schema loss
Prompt Engineers - Multi-table JSON → compact tables

Before (JSON): 1,247 chars - 312 tokens
After (TOON): 489 chars - 122 tokens
Saved: 190 tokens (61%) - Works in any LLM!

Live Stats You Get

Metric	Shows
Bytes	JSON vs TOON (file size)
Tokens	~4 chars = 1 token estimate
Savings	Absolute + % reduction
Real-time	Updates as you type

Token math: (JSON chars - TOON chars) ÷ 4 = tokens saved

Compared to JSON Minify

Format	Size	LLM Friendly?	Readable?	Schema?
JSON (pretty)	100%	❌ Verbose	✅	✅
JSON (min)	65%	❌ Still quotes	❌	✅
TOON	35%	✅ Perfect	✅	✅

Example: 100 Users Database

JSON (12KB → 3K tokens):

[{"id":1,"email":"user1@ex.com","plan":"pro","score":85},{"id":2,...}]

TOON (4.2KB → 1K tokens):

Monthly savings: $10-50+ (1000 API calls)

Perfect For Your Stack

ChatGPT + Claude + Gemini + Llama
× Next.js APIs × Supabase × Firebase
× Zapier × Make.com × n8n

Copy-paste ready for system/user messages.

Bonus Features

Offline - No server calls
Example data - 1-click load
JSON formatter - Pretty print input
Mobile-first - Touch-friendly
Dark mode ready - Matches your theme

Try it now ↑ - Paste any JSON and watch tokens disappear!