Hi! question/doubt, suppose the response is fully streamed, lets say at the end the response will be about 2000 tokens, but of course, it deliver the tokens from the start. The question is, if i cut the streaming, the API will still consume the 2000 tokens? (if i cut the transmition at 600 tokens for example, the response tokens will still be 2000 at the bill?) maybe a basic question but deep doubt! because with this, we could have a fine control over billing/usage against users
#Stream billing
exotic yacht