logoalt Hacker News

nylonstrungtoday at 6:59 PM2 repliesview on HN

I don't really understand how async tool calls translate to token savings

It says that it removes tokens wasted while a model is waiting on synchronous tool calls. What tokens exactly is Pi using when waiting?


Replies

332451btoday at 9:36 PM

"More tool work per model turn" could reduce the number of cache reads (or even cache misses) and associated cost?

dumberquestionstoday at 7:24 PM

I think a running agent periodically checks if a process it started has finished.

show 1 reply