GPT3 - API output quality Vs Playground

Hi guys!

Is anyone using the GPT3 API? I’m kinda struggling with the output quality.
In the playground, it’s excellent, but then in the API Call section, when I initialize the call… the quality is way off compared to the playground.
I’m using the same prompt, etc.

Any ideas?

is the playground using different prompts than you in the api connector?

Nope, exactly the same prompt.
I’m using dynamic command, so .