Gpt-4 + chat/completions slow responce from api

Firstly im a project manager, not a developer. I have some dev knowledge, but I am managing a project using OpenAI and my developers are struggling to complete the task. Perhaps my lack of knowledge is the issue.

I wonder if someone can help;

The requirment is fairly easy to explain. But I have issues my developers seem unable to overcome;

I want to send articles to openai API to request rewriting.

The issue they have encountered is that the responce is too slow to come back. To give an example, if we send a 285 word article, OpenAI is taking over 5 minutes to repond with a rewritten article.

They have tried using;

apiEndpoint: v1 / chat / completions
model: “gpt-4”,

We send the short 250 word article to openai API, and request it to write an article of around 2500 words, but the responce takes over 5 minutes to come back from the API.

Is it normal for gpt-4 to take this long to respond with content of this size?

Cheers guys

5 minutes is pretty slow.

Try using the playground to first confirm that it’s not an issue with the client that you’re using. You should also turn streaming on to determine if it’s a slow token generation, or if for some reason it’s taking a while to start.

thank you. I notice that gpt-4 isnt listed as a model there?

Sorry. That is my fault. I guess my history has it set to “Completions”.

At the top where it says Playground there is a tab to select a different model type. It’s probably on Complete right now and needs to be Chat to see GPT-4

Or just click this: