Why is ChatCompletionService#createStreaming blocking the thread for a significant amount of time #646
Description
When attempting to stream data with a client like so:
try (var stream = client.chat().completions().createStreaming(params))
the thread is blocked at this line of code for 5-15 seconds after which the response is nearly instant
It feels like the data is getting buffered until the entire response is returned and only then will the stream return
Using ChatModel.GPT_5_NANO