What is Triton Inference Server?

Triton Inference Server is NVIDIA's open-source model-serving framework, originally released in 2018, that serves any AI model — LLMs, vision, audio, classical ML — through a unified HTTP/gRPC API with high-throughput batching, multi-model deployment, and rich monitoring. The platform strengthens