Moonshot

Building the foundation for running extra-large language models

Cloudflare demonstrates 3x performance gains for LLM inference by disaggregating prefill and decode compute stages and optimizing KV cache management with prompt caching, enabling efficient multi-GPU scaling on Workers AI.

What I’ve been building: ATOM Report, post-training course, finishing my book, and ongoing research

The Relative Adoption Metric reveals Gemma 4 and Chinese models (Moonshot, Z.ai) gaining traction in the open LLM market, shifting adoption dynamics away from GPT-OSS dominance.

Anthropic data harvested 🤖, AI memo crashes stocks 📉, don't quit your job 💼

Anthropic accuses DeepSeek, Moonshot, and MiniMax of operating 24,000 fraudulent accounts to distill its models, exposing API security vulnerabilities as competitive pressure in AI development intensifies.