search

openai / tiktoken

tiktoken is a fast BPE tokeniser for use with OpenAI's models.
MIT License
11.76k stars 801 forks source link

Cost for failed GPT requests #140

Closed krishna-cactus closed 1 year ago

krishna-cactus commented 1 year ago

Request to Chat models (gpt-3.5-turbo or gpt-4) can fail in 2 different ways:

Issue from OpenAI side - This is case where model is overloaded with other requests

Client adding a request time out - Here, for example, I as customer make a call with request time out as 10 seconds. The moment 10 seconds are passed, request is failed.

So in both situations, am I charged?

hauntsaninja commented 1 year ago

This isn't the appropriate venue for this kind of question, try the following resources: