How streaming LLM APIs work

I decided to have a poke around and see if I could figure out how the HTTP streaming APIs from the various hosted LLM providers actually worked. Here are my notes so far.

Read more here: External Link