Understanding pricing for assistant api threads

Can anyone help me understand how the pricing for assistants api work ?

What I already know:

  • I know that the cost is based on the input and the output tokens, and the assistant threads maintain a history, so every time you will make a request you’ll be charged for all the tokens that have been used (including the history).

What I don’t know:

  • I don’t really know if i’ll be charged for those unwanted threads (that i didn’t delete). If i’m not really using those threads would i still be charged anything ?
  • I want to delete all those unwanted threads, but it seems like openai doesn’t have a native method to list thread and delete those.
  • I’m aware that you can delete thread by the thread id but in the process of testing, i created lots of threads and didn’t save the thread id at all.

Would really appreciate anyone helping me out here. Thanks in advance.

1 Like

you’ll be charged during generation. it does not matter how many threads you currently have. you will not be charged for those. however, if you use File Search and have files stored, then you might be charged for those files ( $0.10 / GB of vector-storage per day (1 GB free)).

1 Like