Different result between API and Playground

I started using the API for mass production of texts, but the contents are getting divergent results instead of following exactly what I put in the prompt. What could be happening?
I’m using the Fine Tuning.

Welcome to the dev community forums!

A handful of people seem to be reporting this, but I’m still skeptical. Are you using chat completions or assistants?

If you use chat completion, you can click on generate code in the top right, (</>) to ensure that your API call has the same configuration as your playground call.

1 Like