This is an n8n community node that lets you encode JSON data into TOON (Token-Oriented Object Notation) format in your n8n workflows.

TOON is a token-efficient alternative to JSON designed specifically for Large Language Model (LLM) prompts. It reduces token usage by 40-60% while remaining human-readable and LLM-friendly.

n8n is a fair-code licensed workflow automation platform.

Available Nodes:

• TOON Encode: Convert JSON to compact TOON format (40-60% token savings)
• TOON Decode: Convert TOON back to JSON for further processing

Table of Contents

• n8n-nodes-toon-encode
  • Installation
  • Features
  • Operations
    • TOON Encode
  • Compatibility
  • Usage
    • Basic Example
    • Use Cases
    • Tips
  • Development & Testing
  • Resources
  • License

Follow the installation guide in the n8n community nodes documentation.

Community Node Name: n8n-nodes-toon-encode

Or install via npm in your n8n installation:

npm install n8n-nodes-toon-encode

✅ Token Savings: Reduce LLM token usage by 40-60%
✅ Human-Readable: Easy to read and understand
✅ LLM-Optimized: Designed for AI/LLM consumption
✅ No Dependencies: Works on n8n Cloud and self-hosted
✅ Flexible Output: Configurable indentation, delimiters, and length markers
✅ Perfect for APIs: Efficiently encode API responses for LLM processing

Encodes JSON data (objects, arrays, or any JSON-serializable data) into TOON format.

Parameters:

Options:

• Minimum n8n version: 1.0.0
• Tested with: n8n 1.x
• Works on: n8n Cloud and self-hosted installations

This node has no external dependencies and is fully compatible with n8n Cloud.

1. Add the TOON Encode node to your workflow
2. Connect it after a node that outputs JSON data
3. The encoded TOON data will be available in the specified output field

Example Input:

{
  "users": [
    {"name": "Alice", "age": 30, "city": "NYC"},
    {"name": "Bob", "age": 25, "city": "LA"}
  ]
}

Example Output (in toon field):

users[2]{name,age,city}:
  Alice,30,NYC
  Bob,25,LA

1. LLM Prompts: Reduce costs by encoding API responses before sending to OpenAI/Anthropic
2. Data Transformation: Convert verbose JSON to compact, readable format
3. Logging: Create human-friendly logs from complex data structures
4. API Integration: Prepare data efficiently for LLM-based processing

• Use ={{ $json }} to encode the entire output from the previous node
• Use ={{ $json.results }} to encode a specific field
• Set Indent to 0 for maximum token savings (compact format)
• Enable Length Markers when passing arrays to LLMs for better reliability

This node is tested against the official TOON test suite to ensure 100% spec compliance.

Run Tests:

npm install
npm run build
npm test

See test/README.md for comprehensive testing documentation.

• n8n community nodes documentation
• TOON Format Documentation
• TOON Specification & Examples

This node includes inlined code from the TOON library by Johann Schopplich, licensed under MIT.

MIT License - See LICENSE file for details.