AIInternal knowledge-base spending reportedly falls roughly 100-fold
A user says their internal knowledge base moved from Opus to Sonnet, then to GLM 5.2 and GLM 5.3 Flash.
TLDR
A user reports that spending on their internal knowledge base was down about two orders of magnitude—roughly 100-fold—since spring 2026. In their September 18 account, they describe moving from Opus to Sonnet to GLM 5.2 to GLM 5.3 Flash.
Combined views
3.9K
2 Sources, first seen 2d ago
54 likes
Internal knowledge-base spending reportedly falls roughly 100-fold
A user says their internal knowledge base moved from Opus to Sonnet, then to GLM 5.2 and GLM 5.3 Flash.