Token Costs

Daily-updated LLM pricing data for OpenAI, Anthropic, Google, and OpenRouter.

What Is This?

An npm package and JSON API that gives you up-to-date token pricing for major LLM providers. Stop hardcoding prices or manually checking pricing pages.

import { CostClient } from 'token-costs';

// Create a client (fetches from remote API)
const client = new CostClient();

// Get pricing for a model
const result = await client.getModelPricing('openai', 'gpt-4o');
console.log(`Input: $${result.pricing.input}/M tokens`);
console.log(`Output: $${result.pricing.output}/M tokens`);

// Calculate cost for an API call
const cost = await client.calculateCost('anthropic', 'claude-sonnet-4', {
  inputTokens: 1500,
  outputTokens: 800,
});
console.log(`Total cost: $${cost.totalCost.toFixed(6)}`);

Or fetch directly without dependencies:

const data = await fetch('https://mikkotikkanen.github.io/token-costs/api/v1/openai.json')
  .then(r => r.json());

Features

• Daily updates - Crawled automatically at 00:01 UTC
• 4 providers - OpenAI, Anthropic, Google, OpenRouter
• Custom providers - Add your own models or override pricing
• Offline mode - Work without network access using custom data
• Zero dependencies - npm package has no runtime dependencies
• TypeScript - Full type definitions included
• Caching - Fetches once per day, caches automatically
• Stale detection - Know when data might be outdated

Installation

npm install token-costs

Custom Providers & Offline Mode

Add custom models or use entirely custom pricing data:

// Add custom models alongside remote data
const client = new CostClient({
  customProviders: {
    'my-company': {
      'internal-llm': { input: 0.50, output: 1.00 }
    },
    'openai': {
      'gpt-4-custom': { input: 25, output: 50 } // Override/add to openai
    }
  }
});

// Offline mode - no remote fetching
const offlineClient = new CostClient({
  offline: true,
  customProviders: {
    'openai': {
      'gpt-4o': { input: 2.5, output: 10 }
    }
  }
});

Documentation

Full usage guide, API reference, and data formats: mikkotikkanen.github.io/token-costs

What's Included

token-costs/
├── CostClient        # Main client class with caching
└── TypeScript types     # Full type definitions

API Endpoints (JSON):

• api/v1/openai.json
• api/v1/anthropic.json
• api/v1/google.json
• api/v1/openrouter.json

Contributing

Found incorrect pricing? Want to add a provider? Contributions welcome!

git clone https://github.com/mikkotikkanen/token-costs
cd token-costs
npm install
npm run build
npm test

See AGENTS.md for development details.

For LLM Providers

We'd prefer not to scrape. Consider publishing /llm_prices.json on your website - a simple standard format that tools can fetch directly. See the full proposal on the documentation site.

License

MIT