From the source
Amp switched to a different model for compacting or summarizing threads, resulting in 4-6x faster performance, roughly 30x lower cost, and improved summary quality.
From the source
Amp now uses a different model when compacting or summarizing threads. It's 4-6x faster, roughly 30x cheaper, and provides better summaries when you either compact a thread or create a new thread with a summary.
ampcode.com
