Open weights consolidate around a flash tier and an Nvidia-owned hub

Open-weight models · Aug 28 – Aug 28

Organizations

Updates

  1. Zhipu's MIT-licensed GLM-5.3-Flash (320B total, 18B active) and Alibaba's Qwen3.8-Flash-Next (6B active, priced about a third below DeepSeek-V4-Flash) both shipped open weights, while Nvidia agreed to buy Hugging Face, the main open-model hub, for $12.9B.

    Zhipu's GLM-5.3-FlashAlibaba's Qwen3.8-Flash-NextNvidia to acquire Hugging Face for $12.9B

Predictions

  1. Feb 2027

    The flash tier (sub-20B active, near-million-token, self-hostable) becomes the default substrate for cost-sensitive agentic work, and control of open-model distribution becomes contested now that a chip vendor owns the hub. Aug 28