How to Use the JSON to TOON Converter

Choose Your Conversion Direction

Click the mode button to switch between JSON → TOON or TOON → JSON conversion. The converter is bi-directional, allowing seamless conversion in both directions.

Input Your Data

Paste your JSON or TOON data into the input editor, or click "Open File" to load from a file. The converter will validate and process your data in real-time.

Customize Options (JSON → TOON)

When converting JSON to TOON, customize the output with these options:

Delimiter: Choose comma (,), tab (\t), or pipe (|) for array values. Tabs often save more tokens.
Indentation: Select 2 or 4 spaces for nested structure indentation.
Show Array Count: Toggle visibility of array length indicators in the output.

Review Token Statistics

When converting to TOON, check the token statistics panel to see how many tokens you're saving. This helps you optimize your LLM prompts for cost efficiency.

Copy or Download the Result

Use the "Copy" button to copy the converted output to your clipboard, or click "Download" to save it as a file for later use.

Why Convert JSON to TOON?

TOON (Token-Oriented Object Notation) is specifically designed to reduce token usage when working with Large Language Models (LLMs), while maintaining human readability and structure. Here are the key benefits:

Save 30-60% on Tokens

TOON typically reduces token count by 30-60% compared to formatted JSON, especially for uniform arrays of objects. This translates to significant cost savings when working with LLMs.

Better LLM Accuracy

TOON's explicit array lengths [N] and field headers help LLMs track structure better, achieving 73.9% accuracy vs JSON's 69.7% in benchmarks.

Human-Readable

TOON maintains readability with YAML-like indentation and CSV-like tabular format. It's easy to read, edit, and understand at a glance.

Perfect for Tabular Data

Excels with uniform arrays of objects - define fields once, then stream data as rows. Ideal for datasets, API responses, and structured data.

Example: Token Savings in Action

JSON (verbose)

{
  "users": [
    { "id": 1, "name": "Alice", "role": "admin" },
    { "id": 2, "name": "Bob", "role": "user" }
  ]
}

TOON (compact)

users[2]{id,name,role}:
  1,Alice,admin
  2,Bob,user

Frequently Asked Questions

What is TOON format?

TOON (Token-Oriented Object Notation) is a compact, human-readable serialization format designed specifically for passing structured data to Large Language Models with significantly reduced token usage. It's a lossless, drop-in representation of JSON data optimized for LLM contexts.

How much can I save with TOON?

TOON typically saves 30-60% tokens compared to formatted JSON, especially for uniform arrays of objects. The actual savings depend on your data structure - tabular data with many repeated keys benefits most, while deeply nested or non-uniform data may see smaller savings.

Which delimiter should I use?

The choice depends on your data and use case:

Comma (,) - Default, familiar, works well for most cases
Tab (\t) - Often saves more tokens, single character, rarely needs escaping
Pipe (|) - Visual clarity, good middle ground between comma and tab

Can I convert TOON back to JSON?

Yes! This converter is bi-directional. Click the mode button to switch to "TOON → JSON" conversion. TOON is designed to be a lossless format, so you can convert back and forth without losing data.

When should I NOT use TOON?

TOON excels with uniform arrays of objects, but there are cases where other formats are better:

Deeply nested or non-uniform structures - JSON-compact often uses fewer tokens
Pure tabular data without nesting - CSV is more compact
Semi-uniform arrays (40-60% tabular eligibility) - Token savings diminish
Latency-critical applications - Benchmark on your specific setup

Is my data safe?

Absolutely! All conversion happens entirely in your browser using client-side JavaScript. Your data never leaves your device, is not sent to any server, and is not stored anywhere. This tool is 100% privacy-focused.

How accurate is the token counter?

The token counter provides an approximation based on typical GPT-style tokenization patterns. Actual token counts may vary slightly depending on the specific tokenizer used by your LLM (GPT-4, Claude, etc.). Use it as a guide for relative savings rather than exact counts.

Can I use TOON in production?

Yes! TOON is production-ready and follows a stable specification (v2.0). However, it's designed primarily for LLM input/output, not as a replacement for JSON in APIs or storage. Use JSON programmatically and convert to TOON for LLM contexts.

Free Powerful JSON ↔ TOON Converter

JSON Input

TOON Output