Only for AI Variables
When using AI variables for content generation, it is important to note that the number of credits consumed is an estimate and not a fixed value. This is due to how costs are calculated, which are based on the total number of tokens, and these are not always predictable.
💰 Why can't the exact cost be calculated?
Tokens from referenced variables: The variables you input, marked with
{}and highlighted, may contain a variable number of tokens depending on the text they include.Tokens from the generated response: The number of tokens used to create the response is also uncertain until the content generation is complete.
Therefore, the final credit cost is defined only at the time of execution, as it is not possible to determine the exact number of tokens used beforehand.
What are tokens?
Tokens are pieces of text. They can be as small as a single letter or as long as a complete word, depending on how they are divided. The concept of tokens is used to process and calculate prompt costs. For example:
The word "intelligence" could be a single token.
However, a more complex word or one with special characters could be split into multiple tokens.
A simple example:
"Hello, world" → 3 tokens:
Hello,,andworld.
How are tokens calculated?
The model divides both the input text (prompt) and the generated output (response) into tokens for processing. The total cost is based on:
Tokens from the prompt: This is the content you write or define as input.
Tokens from the response: This is the content generated as output.
For example:
If the prompt has 50 tokens and the generated response has 100 tokens, the total will be 150 tokens.
💡 Key considerations:
Credits as an estimate: The system calculates credits based on the total tokens. However, since the exact number of tokens in the variables or the generated response cannot be predicted, the value shown beforehand is approximate.
Final execution: Only at the moment of execution can the precise total cost be calculated based on the tokens actually used.