'Which Cheap Chinese LLM Are You Using?' — HN Signals a Cost-Driven Shift to Open Weights

Hacker News·low signal

A Hacker News thread asking which inexpensive Chinese models developers actually run captures a real adoption trend: builders are routing everyday tasks to Qwen, DeepSeek, GLM, Kimi, and MiniMax rather than premium frontier APIs. The cost gap, combined with open weights now dominating Hugging Face's trending charts, is reshaping default tooling choices for solo developers and small teams.