Tokens

A Token is essentially a "word".

As part of your subscription you have a limited number of tokens each month. RAIA uses OpenAI GPT 4 as the primary model (unless you request otherwise). All tokens are measured via OpenAI and if it reaches your limit each month, you will have the option to turn off your Agent or pay usage fees.

RAIABot.com Token Pricing (as of Mar 3, 2024)

Program	Number Tokens Included / Month	Overage Costs
Out of the Box	100,000	$1 per 20,000 Tokens
Professional	500,000	$1 per 20,000 Tokens
Small Business	1,000,000	$1 per 20,000 Tokens

Approximately, 1,000,000 tokens equate to about 750,000 words in English.

What is a Token?

OpenAI uses a concept called "tokens" to measure and bill for the usage of its services, like ChatGPT. Imagine tokens as pieces of digital currency, where each piece represents a certain amount of text processed by OpenAI's models.

Here's a simple breakdown:

1. What is a Token? A token is a small unit of text, which could be a word or part of a word, punctuation (like "." or ","), or space. The definition of a token is based on how text is split and understood by the AI model.

2. How Tokens Are Used for Billing: Whenever you use OpenAI's services to generate text (like asking a question or creating content) or to have the AI understand and respond to text you provide, that text is measured in tokens. The total number of tokens processed determines how much you are billed.

3. Why Tokens? Tokens provide a way to measure the work the AI does in a consistent manner, whether it's understanding a question you've asked or generating a response. This is because longer texts require more computational power and resources to process.

In the simplest terms, think of tokens as slices of bread. If you want to make a sandwich (generate some text), each slice (token) contributes to the size of the sandwich. The bigger the sandwich (the more text there is), the more slices (tokens) you'll need. And just like buying bread, more slices mean a higher price.