toon-format/toon

🎒 Token-Oriented Object Notation (TOON) – Compact, human-readable, schema-aware JSON for LLM prompts. Spec, benchmarks, TypeScript SDK.

67
/ 100
Established

TOON helps developers working with Large Language Models (LLMs) by providing a more compact and readable way to format structured data inputs. It takes existing JSON data and translates it into a token-efficient format that LLMs can process reliably, reducing both cost and potential errors. This is primarily useful for AI/ML engineers and data scientists preparing data for LLM prompts.

23,207 stars. Used by 21 other packages. Actively maintained with 1 commit in the last 30 days. Available on npm.

Use this if you need to provide structured data, especially uniform lists or tables, to Large Language Models while minimizing token usage and improving the model's ability to parse information accurately.

Not ideal if your data is extremely complex with many nested levels, if you're working with purely tabular data where CSV is already sufficient, or if ultra-low latency is your absolute highest priority.

LLM-prompting AI-data-preparation token-optimization structured-data natural-language-processing
Maintenance 13 / 25
Adoption 15 / 25
Maturity 22 / 25
Community 17 / 25

How are scores calculated?

Stars

23,207

Forks

1,033

Language

TypeScript

License

MIT

Last pushed

Mar 10, 2026

Commits (30d)

1

Reverse dependents

21

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/toon-format/toon"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.