Free API Pricing Data — LLM Prices JSON Download

Download the dataset

GET https://apicostcalc.com/pricing.json Full dataset — 387 LLM models + 34 API prices · 88KB · Updated monthly

curl https://apicostcalc.com/pricing.json

The file is a plain JSON with two top-level arrays: llm_models (LLM pricing per 1M tokens) and api_prices (provider service pricing). No auth required.

Data format

LLM models (`llm_models` array)

{
  "id": "openai/gpt-4o",
  "name": "GPT-4o",
  "provider": "OpenAI",
  "input_per_1m_usd": 2.50,
  "output_per_1m_usd": 10.00,
  "context_k": 128
}

Field	Type	Description
`id`	string	Slug: `provider/model-name`
`name`	string	Model display name
`provider`	string	Provider name (OpenAI, Anthropic, etc.)
`input_per_1m_usd`	number	USD cost per 1M input tokens
`output_per_1m_usd`	number	USD cost per 1M output tokens
`context_k`	number	Context window in thousands of tokens

API provider prices (`api_prices` array)

{
  "provider": "Stripe",
  "service": "card_processing",
  "price": "3.1% + $0.30",
  "metric": "per successful card charge",
  "free_tier": "",
  "last_verified": "2026-07-15"
}

Coverage

Category	Providers covered	Count
LLM / AI models	OpenAI, Anthropic, Google, Mistral, Meta, DeepSeek, xAI, Cohere, Together AI, Fireworks + 20 more	387 models
Payments	Stripe, Paddle, PayPal, Adyen, Braintree	8 prices
Email API	SendGrid, Resend, Postmark, SES, Mailgun	7 prices
SMS / Voice	Twilio, Vonage	4 prices
Maps / Location	Google Maps, HERE, MapBox	4 prices
Search / Data	Algolia, Brave Search, Bing Search, Alpha Vantage, CoinGecko	6 prices
Other	ElevenLabs, Replicate, Deepgram, AssemblyAI and more	5+ prices

License & attribution

The data is available under Creative Commons BY 4.0. You can use it in products, research, and commercial applications. Required attribution:

Data source: APICostCalc.com

A link back to apicostcalc.com is appreciated but not required for machine use.

What you can do

Use in your own product or SaaS
Include in open-source projects
Build npm/Python packages around it
Use in research and academic work
Redistribute with attribution

Data quality notes

LLM prices updated monthly from official provider pricing pages
Prices are standard public rates — enterprise pricing not included
Some providers have complex tiered pricing; we show the base/starter tier
Found an error? Report it →

Usage examples

Find the cheapest model for a workload

import json, urllib.request

with urllib.request.urlopen('https://apicostcalc.com/pricing.json') as r:
    data = json.loads(r.read())

# Cost for 1M input + 200k output tokens
models = data['llm_models']
results = [
    (m['name'], m['provider'],
     m['input_per_1m_usd'] * 1 + m['output_per_1m_usd'] * 0.2)
    for m in models if m['input_per_1m_usd'] > 0
]
results.sort(key=lambda x: x[2])
for name, provider, cost in results[:5]:
    print(f"{provider}/{name}: ${cost:.4f}")

JavaScript / Node

const res = await fetch('https://apicostcalc.com/pricing.json');
const { llm_models } = await res.json();
const gpt4o = llm_models.find(m => m.id === 'openai/gpt-4o');
console.log(`GPT-4o input: $${gpt4o.input_per_1m_usd}/1M tokens`);

Need more?

The free dataset is updated monthly and covers the most common providers. If you need:

Webhooks when prices change
Price history (track how prices changed over time)
Regional pricing differences
Real-time updates
Custom provider coverage