TOON

Tracked

A token-oriented, compact and schema-aware data notation for LLM prompts and serialization.

Author toon-format Open Sourced 2025-10-22 Last Commit Unknown

TOON (Token-Oriented Object Notation) is a compact, human-readable data format designed as a token-efficient alternative to JSON for LLM prompts and structured data serialization. It uses explicit token delimiters and schema-aware parsing to reduce prompt size while maintaining readability, making it particularly effective for passing structured data to and from large language models within limited context windows.

Format and Schema

  • Compact delimiters and optional schema validation achieve significant token savings over JSON
  • Pattern-based schemas for data validation and backward compatibility across evolving data structures
  • Explicit token-splitting rules and lightweight semantic conventions enable deterministic parsing without ambiguity
  • Balances human readability, machine verifiability, and minimal token consumption within LLM context windows

Toolchain and Integration

  • TypeScript SDK with parsers, serializers, and schema validators for both frontend and backend integration
  • Benchmarking tools to measure token efficiency gains in real workflows
  • Straightforward adoption path with comprehensive documentation and examples
  • Designed as a lightweight serialization format for small structured payloads between services

Prompt Engineering Applications

  • Prompt engineers craft concise structured inputs for LLM calls, reducing token costs
  • Teams building prompt template libraries adopt TOON as a standardized interchange format for reusable components
  • Services exchange small structured payloads where token efficiency matters
  • Compatible with existing LLM workflows as a drop-in JSON replacement