TOON to JSON Converter

TOON → JSON converter

TOON 114 tokens
JSON 321 tokens
Going to JSON adds +182% +207 tokens / call

At calls/day on sending this as JSON instead of TOON costs about $1,889 more per year in input tokens (at $2.50 / 1M input tokens — estimate)

JSON 1284 chars vs TOON 456 chars (64% smaller). Token counts are estimates (~4 chars/token) compared with compact JSON — confirm exact counts with your model's tokenizer. Savings are largest for arrays of uniform objects.

TOON (Token-Oriented Object Notation) is great for shrinking the structured data you send to an LLM, but almost everything downstream — APIs, databases, validators, your own code — expects JSON. This converter does the reverse trip: paste TOON and get back clean, pretty-printed JSON instantly, with values restored to their proper types (numbers, booleans, null, and strings).

It handles the full format: objects, scalar arrays, the tabular header-plus-rows form used for arrays of uniform objects, nested structures, and quoted strings — across comma, tab, and pipe delimiters. Everything runs in your browser, so your data never leaves your device. Need to go the other way and compress JSON for a prompt? Use the JSON → TOON converter.

How to use this converter

  1. Paste your TOON into the left box.
  2. Pick the matching delimiter (comma, tab, or pipe) — it must match what the TOON uses, or rows will misparse.
  3. Read the JSON on the right — it updates as you type. Invalid TOON shows a parse error instead.
  4. Copy the JSON and use it in your code or tooling.

TOON to JSON — FAQ

What does this converter do?
It parses TOON (Token-Oriented Object Notation) and turns it back into standard, pretty-printed JSON. Paste TOON on the left and get valid JSON on the right — useful when a model returns TOON, or when you have TOON-formatted data you need to feed into JSON tooling.
What is TOON?
TOON (Token-Oriented Object Notation) is a compact, indentation-based data format designed to use fewer tokens than JSON when sent to an LLM. Arrays of uniform objects become a tabular header plus rows instead of repeating every key, which is where most of the token savings come from.
Will the JSON exactly match the original?
For data produced by a faithful TOON encoder, yes — this decoder round-trips objects, scalar arrays, tabular arrays of objects, nested structures, and quoted strings. Values are restored to their JSON types (numbers, booleans, null, strings). Hand-written TOON with unusual spacing may need tidying first.
Which delimiter should I select?
Match the delimiter the TOON actually uses — comma, tab, or pipe. The header and rows are split on that character, so choosing the wrong one will misparse rows that contain it. Comma is the most common default; tab is common in token-optimized TOON because it rarely needs quoting.
Does my data get uploaded anywhere?
No. Conversion runs entirely in your browser with JavaScript — your TOON never leaves your device, so it's safe to paste sensitive data.
What if my TOON is invalid?
The converter shows a parse error and clears the output. The most common causes are the wrong delimiter selected, inconsistent indentation (TOON uses two spaces per level), or a row whose column count doesn't match the tabular header.
Why convert back to JSON at all?
TOON is great for shrinking what you send to a model, but most tools, APIs, databases, and validators speak JSON. Converting the model's TOON output (or a TOON snippet) back to JSON lets you parse, validate, and store it with standard libraries.
Is this converter free?
Yes, completely free, no sign-up, and it runs client-side.