DesktopGPT uses OpenAI’s services to bring AI models to your desktop. Each AI model has a different value assigned to it for the cost that is incurred with each prompt. Currently, three the following models are supported by DesktopGPT:
GPT-4
GPT-3.5 4 Turbo
GPT-44o
GPT-4o-4 Turbomini
When using each model, a token is weighted against the cost that is incurred when prompting ChatGPT with the following multipliers:
GPT-3.5 Turbo, 1X (Deprecated)
GPT-4 Turbo, 2.5X
GPT-4, 5X
GPT-4o, 0.33X
GPT-4o-mini, 0.33X
Meaning, that when using GPT 4, it is 5x the token cost of GPT 3.5 Turbo. As an example, when using 3.5 Turbo, a prompt that uses 2000 tokens would use 10000 tokens to ask the same question as 3.5 Turbo.
For The GPT-4o models are currently the most cost efficient and for most users , 3.5 Turbo will provide sufficient results, we . We provide model options so that a user can compare and contrast outputs if they are not satisfied with the base model.
...
Stardock is evaluating token usage during the beta period and will frequently update these weights to better reflect usage of the API.
Available Account Tokens and Purchasing
Options for viewing available DesktopGPT tokens and purchasing new ones can be found through the DesktopGPT UI’s Account tab.