Embedded dev on Chinese models: DeepSeek and GLM are good enough, skip Codex/Claude

sven_ai · x · 2026-09-29

A post that drew nearly 800K views: an embedded C++ / Python testing developer says he only uses DeepSeek and GLM, since his work isn't complex enough to need Codex or Claude — Chinese models have yet to hit a problem they can't solve. The thread sparked debate on whether premium coding models are overkill for ordinary engineering work.

Original post →

More from coding & agent

coding & agent channel →