streamCompletion

Calls the completions API endpoint and streams each token 1 at a time for a faster response time.

This method is technically not blocking, but the returned iterable will block until the next token is generated.

// Each iteration of the loop will block until the next token is streamed
for (chunk in openAI.streamCompletion(request)) {
    // Do something with the chunk
}

Completions are considered Legacy, and OpenAI officially recommends that all developers use the chat completion endpoint isntead. See streamChatCompletion.

Parameters