“If you come back after one hour and one minute, that cache is gone… cache writes are expensive… leaving sessions idle for more than an hour is expensive.”
DS
Dave Slutzkin
The Startup Podcast ·
“You just compact manually and you save a significant amount of money, even at that point, because it doesn't do a cache write”
Dave Slutzkin · The Startup Podcast · Applied Wizardry: Cutting AI coding token costs, with Dave Slutzkin
Play from 19:42
19:42
Audio streams from the publisher’s own feed. Verbatim does not host or copy it.
What was described
After being idle for more than an hour, manually trigger a server-side compaction instead of letting the session write a cache, then continue the session
Said nearby on the same subject
“We learned that when you took an AI recommended agenda, you were 30% more likely to actually show up to your sessions than when you built your agenda yourself.”
♫
Said on The CMO Podcast
“I have my auto-compact window in Anthropic models set to about 300,000 or 350,000 tokens, but I will compact manually over about 200,000.”
DS
Dave Slutzkin
The Startup Podcast ·
“I wasn't sure if I would be able to continue. But I wrote it out. I was like, I'm really passionate about this. What can I do to keep myself busy and help people? So I started doing these short training, recorded training sessions.”
RS
Risa Seelenfreund
Career Change & Business Advice for Women | Second Act Success Podcast ·
Moments like this, every Monday
The strongest new moments on your topics, in the speakers' own words. Free for three topics.
No card. No password. One-click unsubscribe.
Track a competitor?
Pro watches up to 25 names or phrases and emails you the morning after they are said.
See Pro, $29 a monthVerbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.