Calculate the exact cost of your AI API usage, not a vague subscription guess.
Enter your input/output tokens per request and request volume. AIStackCost multiplies against official per-token prices for GPT-5.6, Claude, Gemini, Grok and DeepSeek so you get a real daily, monthly and annual number — not a placeholder range.
Official per-token pricesCSV exportPrintableNo signup
| Model | Provider | Input $/1M | Output $/1M | Cost/request | Daily | Monthly | Annual | Source |
|---|
Prices are approximate planning estimates. Rates shown are standard per-token API prices without prompt caching, batch discounts, or long-context surcharges, verified against each provider's official pricing page on 2026-08-18. Providers change prices without notice — confirm the current rate at the official source before budgeting.
Why per-token, not per-seat
A subscription price tells you what a human costs to use a chat app. API pricing tells you what your product costs to run per request — the number that actually matters once you're calling a model in code.
Caching and batch cut cost hard
Repeated system prompts or few-shot context can be cached for a fraction of standard input price on most providers, and batch/async processing typically halves cost again. Check the official page for exact rates once you pick a model.
Output tokens dominate cost
Output tokens cost 3-6x input tokens on almost every model here. A prompt that asks for a short, structured answer instead of a long one is often the single biggest lever on your bill.
AI cost guides
Related planning guides for choosing and budgeting an AI stack.
Are these AI API prices guaranteed?
No. Prices are entered from each provider's official API pricing page and checked periodically, but providers change rates without notice. Treat every result as a planning estimate and confirm the current rate at the official source linked next to each model before budgeting.
What is the difference between API pricing and a ChatGPT/Claude subscription?
A consumer subscription (ChatGPT Plus, Claude Pro, Gemini Advanced) is a flat monthly fee for the chat app with usage limits. API pricing is pay-per-token: you're billed for every request your product or script sends, no flat fee. This calculator is for API/developer usage, not consumer subscriptions.
Does this include prompt caching or batch API discounts?
No, the base calculation uses standard per-token rates. Prompt caching and batch processing can cut real cost by 50-90% depending on the provider and how repetitive your prompts are. Check the official pricing page linked for each model for exact rates before finalizing a budget.
How do I estimate tokens from text?
Use the token estimator above: paste a sample prompt or response and it estimates tokens at roughly 4 characters per token for English text. This is a rough approximation; exact counts depend on the model's tokenizer.
Is my data uploaded?
No. The calculation and exports run in your browser. Do not paste sensitive or private text into the token estimator.