NVIDIA NIM

NVIDIA NIM (NVIDIA Inference Microservices) is a platform for deploying optimised AI models as containerised microservices on NVIDIA GPUs. It packages foundation models with NVIDIA's TensorRT-LLM optimisation for maximum inference throughput. NIM provides ready-to-deploy containers for LLMs, vision models, and multimodal models from NVIDIA and partners. Used by enterprises and cloud providers deploying AI inference at scale who want optimised model serving on NVIDIA infrastructure with enterprise support and pre-built containers for rapid deployment.