This suite of cloud-native microservices is meant to streamline the deployment of AI versions across many platforms. NIM delivers optimized inference with a spectrum of AI versions, supporting tasks ranging from facts processing to massive language model customization, inference, retrieval-augmented era, and applying guardrails in AI products.