The Embeddings API returns a vector - a list of numbers like this:
[
-0.006929283495992422,
-0.005336422007530928,
...
-4.547132266452536e-05,
-0.024047505110502243
],
It measures the relatedness of strings. You can use it to implement services like recommendation engines, search, and classification.
Fine-tuning returns text bot response improved by providing examples.
I recommend this post to learn more about embeddings : Introducing Text and Code Embeddings