LiteLLM is moving to Rust. 🦀 Sub-1ms overhead. A sub-100MB binary. The same Python SDK and AI gateway you already use. Here's why and what's changing 🧵 docs.litellm.ai/blog/litellm-r…