ovg-project / kvcached

Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond

kvcache-ai / AgentENV

AgentENV (AENV) is a distributed platform for running agent environments at scale.

kvcache-ai / Mooncake

Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.

kvcache-ai / ktransformers

A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations

uccl-project / uccl

UCCL is an efficient communication library for GPUs, covering collectives, P2P (e.g., KV cache transfer, RL weight transfer), and EP (e.g., GPU-driven)

Zefan-Cai / R-KV

[Neurips 2025] R-KV: Redundancy-aware KV Cache Compression for Reasoning Models

jjang-ai / vmlx

vMLX - Use MLX models easily - JANGQ (GGUF for MLX) - Not dependant on mlx_vlm