Cloudflare shrinks AI models to serve more users for less cost

Original: Smaller, faster, safer: running Kimi and GLM at scale

Source: blog.cloudflare.com ↗