import { encode } from '@toon-format/toon';
// Real-world example: E-commerce product catalog
const ecommerceData = {
products: [
{
id: 1,
name: "Wireless Bluetooth Headphones",
price: 79.99,
category: "Electronics",
in_stock: true,
rating: 4.5,
reviews: 128
},
{
id: 2,
name: "Organic Cotton T-Shirt",
price: 24.99,
category: "Clothing",
in_stock: true,
rating: 4.2,
reviews: 89
},
{
id: 3,
name: "Stainless Steel Water Bottle",
price: 34.99,
category: "Home & Kitchen",
in_stock: false,
rating: 4.8,
reviews: 256
}
],
metadata: {
total_products: 3,
page: 1,
per_page: 10,
last_updated: "2025-12-04T15:30:00Z"
}
};
// Simulate larger dataset (10x more products)
const largeDataset = {
...ecommerceData,
products: Array.from({ length: 30 }, (_, i) => ({
id: i + 1,
name: `Product ${i + 1}`,
price: Math.random() * 100 + 10,
category: ["Electronics", "Clothing", "Home", "Sports", "Books"][Math.floor(Math.random() * 5)],
in_stock: Math.random() > 0.3,
rating: Math.round((Math.random() * 2 + 3) * 10) / 10,
reviews: Math.floor(Math.random() * 500)
}))
};
console.log('=== AI Token Usage Analysis ===\n');
function analyzeData(name, data) {
const jsonStr = JSON.stringify(data);
const toonStr = encode(data);
const jsonTokens = jsonStr.length; // Approximate token count
const toonTokens = toonStr.length;
const savings = jsonTokens - toonTokens;
const savingsPercent = ((savings / jsonTokens) * 100).toFixed(1);
console.log(`${name}:`);
console.log(` JSON size: ${jsonTokens} characters`);
console.log(` TOON size: ${toonTokens} characters`);
console.log(` Savings: ${savings} characters (${savingsPercent}%)`);
console.log(` Cost savings: ~$${(savings * 0.00003).toFixed(4)} per API call`);
console.log('');
return { jsonTokens, toonTokens, savings, savingsPercent };
}
// Analyze different scenarios
const small = analyzeData("Small Dataset (3 products)", ecommerceData);
const large = analyzeData("Large Dataset (30 products)", largeDataset);
console.log('=== Real-World Scenarios ===\n');
// Scenario 1: Daily API calls
console.log('Daily API Usage (1000 calls with large dataset):');
const dailySavings = large.savings * 1000;
console.log(` Token savings: ${dailySavings.toLocaleString()} characters/day`);
console.log(` Cost savings: $${(dailySavings * 0.00003).toFixed(2)}/day`);
console.log(` Monthly savings: $${(dailySavings * 0.00003 * 30).toFixed(2)}`);
console.log('');
// Scenario 2: Chatbot with product data
console.log('Chatbot Response (including product catalog):');
const chatbotTokens = large.jsonTokens;
const chatbotToonTokens = large.toonTokens;
console.log(` JSON approach: ${chatbotTokens} tokens`);
console.log(` TOON approach: ${chatbotToonTokens} tokens`);
console.log(` Faster response with ${large.savingsPercent}% less data`);
console.log('');
// Scenario 3: RAG system
console.log('RAG System (Retrieval-Augmented Generation):');
console.log(` Can fit ${Math.floor(100 / (100 - parseFloat(large.savingsPercent)))}x more data in context`);
console.log(` Better retrieval with structured TOON format`);
console.log(` Reduced latency due to smaller payload`);
