OpenAI’s API operates on a pay-as-you-go model, charging users based on the number of tokens processed. Here’s a breakdown of the costs associated with different models:
GPT-4o:
-
Input (Prompt) Tokens: $2.50 per million tokensWikipedia+6Wikipedia, la enciclopedia libre+6OpenAI Blog – Stephen Walther on OpenAI+6
-
Output (Completion) Tokens: $10.00 per million tokensOpenAI Blog – Stephen Walther on OpenAI+1OpenAI Help Center+1
GPT-4o Mini:
-
Input Tokens: $0.15 per million tokensOpenAI Developer Community+11OpenAI Blog – Stephen Walther on OpenAI+11Wikipedia+11
-
Output Tokens: $0.60 per million tokensOpenAI Blog – Stephen Walther on OpenAI+1Wikipedia+1
GPT-4.5:
-
Input Tokens: $75.00 per million tokensOpenAI Developer Community+8Wikipedia, la enciclopedia libre+8Wikipédia, l’encyclopédie libre+8
-
Output Tokens: $150.00 per million tokensWikipedia
For instance, with GPT-4o, processing 1 million input tokens costs $2.50, and generating 1 million output tokens costs $10.00. Therefore, a $10 expenditure would allow for approximately 4 million input tokens or 1 million output tokens, depending on your usage pattern.Wikipedia+2Wikipedia+2Wikipedia+2
It’s important to monitor your token usage to manage costs effectively. OpenAI provides tools to set usage limits and track expenses, helping you stay within your budget.Reddit+1OpenAI Platform+1
For detailed and up-to-date pricing information, please refer to OpenAI’s official pricing page.
Reviews
There are no reviews yet.