NVIDIA Dynamo

Cloud-native control plane for scalable GenAI inference

Open-source Kubernetes-native building blocks for deploying, routing and scaling GenAI inference, including an LLM gateway, autoscaling, LoRA management and KV-cache offloading.

Kubernetes-native distributed LLM inference stack

llm-d is an open-source Kubernetes-native stack for distributed LLM inference with cache-aware routing and disaggregated serving. It separates prefill and decode stages across different GPU pools for optimal resource utilization, routes requests to nodes with warm KV caches, and integrates with vLLM as the serving engine. Apache-2.0 licensed with 2,900+ GitHub stars.

Kubernetes-native model inference platform

KServe is an open-source Kubernetes-native platform for deploying and managing ML model inference at scale. It provides standardized inference protocols, autoscaling including scale-to-zero, canary rollouts, A/B testing, and multi-model serving. KServe supports all major ML frameworks including TensorFlow, PyTorch, scikit-learn, XGBoost, and LLM runtimes like vLLM and Triton through pluggable serving runtimes.

Fast serving framework for LLMs and vision models

SGLang is an open-source serving framework for large language and vision-language models, designed for low latency and high throughput. It features RadixAttention for automatic KV cache reuse, compressed finite state machines for fast structured output generation, continuous batching, and tensor parallelism. With over 25,000 GitHub stars, it supports models like LLaMA, Mistral, Qwen, and Gemma on NVIDIA and AMD GPUs.

NVIDIA's LLM inference optimization and acceleration library

TensorRT-LLM is NVIDIA's open-source library for optimizing LLM inference on NVIDIA GPUs. It provides kernel fusion, quantization (FP8, INT4, INT8), KV cache optimization, and in-flight batching to maximize throughput. Supports multi-GPU and multi-node setups with tensor and pipeline parallelism, and integrates with Triton Inference Server for production deployment of models like LLaMA, GPT, Mistral, and Qwen.

High-throughput LLM serving engine

vLLM is an Apache-2.0 LLM inference and serving engine focused on high-throughput self-hosted model APIs. It combines PagedAttention, continuous batching, prefix caching, quantization options, OpenAI-compatible serving, structured outputs, metrics, Docker/Kubernetes deployment guidance and integrations with agent and LLM frameworks.

Heterogeneous CPU-GPU inference and SFT for large MoE models

Open-source framework for running and fine-tuning large Mixture-of-Experts models with heterogeneous CPU-GPU execution, optimized kernels, limited VRAM and SGLang or LLaMA-Factory integrations.

Official Kubernetes and Helm reference stack built on the vLLM inference engine

Official vLLM reference implementation for scaling the existing inference engine on Kubernetes with Helm, request routing, KV-cache offload, autoscaling and Prometheus/Grafana observability.

Open-source GPU control plane for scalable AI model serving

Open-source GPU cluster manager that configures vLLM, SGLang, TensorRT-LLM or custom engines, serves models through compatible APIs, and provisions SSH-accessible GPU instances across on-premises, Kubernetes and cloud environments.

Disaggregated KV cache storage and transfer for LLM serving

Open-source infrastructure for disaggregated LLM serving that pools KV caches across prefill and decode workers, with high-performance transfer, distributed storage and integrations for vLLM and SGLang.

Reusable KV cache infrastructure for scalable LLM inference

Open-source KV cache management layer that persists, offloads and reuses model key-value caches across requests and serving engines to reduce repeated prefill work and improve inference throughput.

Quantize and compress LLMs for optimized vLLM deployment

Open-source, Hugging Face-compatible library for quantizing and compressing LLMs (W8A8, W4A16, NVFP4/MXFP4, KV-cache and attention quantization, expert pruning) into vLLM-ready `compressed-tensors` checkpoints.