Sentdex has run 4B+ tokens locally on GLM 5.3 Flash — his most-used local model ever

Sentdex · x · 2026-09-29

ML YouTuber Sentdex reports he's processed over 4 billion tokens locally with GLM 5.3 Flash, calling it by far his most-used local model ever — a strong real-usage endorsement for local inference.

Original post →

More from Infra

Infra channel →