Welcome Back

Enter your details to access your account.

OR
Developer Tools

AI Cost Calculator

Estimate the monthly cost of AI APIs from OpenAI, Anthropic (Claude) and Google (Gemini) from your token usage, or enter your own prices.

Free · No sign‑up Added Jun 2026
~750 words = 1k tokens
Pricing (per 1M tokens):
Prices were checked against each provider's pricing page on November 5, 2026. Providers change prices often, so confirm on the official page before you budget, or use the custom option with your own prices.

About AI Cost Calculator

AI models are charged by usage, and the unit is the token, a small piece of text. As a rough guide, about 750 words come to around 1,000 tokens. Providers charge separately for the text you send in (input tokens) and the text the model writes back (output tokens), and they quote prices per million tokens. This calculator turns your expected monthly token use into a monthly cost.

Pick a model, then enter your monthly input tokens and output tokens. The tool multiplies each by that model's price per million tokens and adds them together. The list covers Anthropic Claude Opus 5.5, Sonnet 5.5 and Haiku 5.5, OpenAI GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra and GPT-6 Luna, and Google Gemini 4 Argon and Gemini 3.8 Flash. The input and output price in dollars per million tokens is shown above the button, and an option at the bottom of the list lets you type your own prices for any other model.

An important limit: the prices are entered in the page, not read live from the providers. They were checked against each provider's pricing page on November 5, 2026, and the page says so. Two announced increases are built in and switch on by date: Gemini 4 Argon rises on December 1, 2026 and Gemini 3.8 Flash on January 1, 2027. Providers change prices and release models often, so confirm on the official page before you budget, or use the custom option. The tool uses standard prices only. It does not include batch discounts, cached-input pricing, long-prompt surcharges, image or audio use, or fine-tuning costs.

Who uses it and when

Budgeting an AI feature

Estimate the monthly cost of a chatbot or tool before you build it.

Comparing models

See how much cheaper a smaller model is for the same amount of text.

Checking a bill

Compare your token use with what you were charged.

Planning for growth

Increase the tokens to see the cost if usage grows.

Why use this AI Cost Calculator

Compare what the same usage costs on different models, with dated prices.

Input and output priced separately

Each is multiplied by its own price per million tokens, as providers charge.

Nine current models and a custom option

Claude 5.5 models, GPT-6 models and Gemini models, or type your own prices for any other model.

Prices shown

The price per million tokens is shown for the model you pick.

How it works

Choose a model

Pick a model from the list, or choose the custom option and type your own input and output prices.

Enter input tokens

Type the number of tokens you expect to send each month.

Enter output tokens

Type the number of tokens you expect the model to write each month.

Calculate

Click the button to see the estimated monthly cost.

Frequently asked questions

What is a token?

A token is a small piece of text that AI models read and write. As a rough guide, 1,000 tokens is about 750 words in English.

How is the cost calculated?

Input tokens divided by one million, times the model's input price, plus output tokens divided by one million, times the model's output price.

Are the prices current?

They were checked against the providers' pricing pages on November 5, 2026. Prices change often, and two increases are already announced for Gemini models, which the tool applies from their start dates. Always confirm on the official page, or use the custom option with your own figures.

My model is not in the list. What should I do?

Choose the custom option and type its input and output price per million tokens from the provider's pricing page. The calculation is the same.

Why are input and output priced differently?

Providers charge more for the text a model generates than for the text it reads, so the tool prices them separately.

What is not included?

Batch and cached-input pricing, surcharges for very long prompts, image and audio use, and fine-tuning. Providers often offer a discount for batch jobs, so your real bill may be lower.

How many tokens is my text?

About 750 English words is roughly 1,000 tokens. Other languages and code can use more tokens for the same length.

Is what I enter stored?

No. The calculation runs in your browser and nothing you enter is uploaded or saved.