Alibaba

RTP-LLM: Alibaba's high-performance LLM inference engine for diverse applications.

End-to-end benchmark for AI-generated GPU kernels, drawn from real production traces — turn a PyTorch reference into a DSL kernel (Triton, Gluon, FlyDSL, CuteDSL) and grade it on compilation, numerical correctness, and speed-of-light efficiency, on both AMD and NVIDIA GPUs.

Alibaba Cloud's high-performance KVCache system for LLM inference, with components for global cache management, inference simulation(HiSim), and more.

An Efficient and User-Friendly Scaling Library for Reinforcement Learning with Large Language Models

A construction kit for reinforcement learning environment management.

Machine learning platform for Web developers

Evidence-based guardrails for AI agents and automated workflows. Versioned policies, structured evidence, and auditable allow/deny decisions.

MNN: A blazing-fast, lightweight inference engine battle-tested by Alibaba, powering high-performance on-device LLMs and Edge AI.

An end-to-end agent project for GPU kernel implementation, analysis, profiling, and iterative optimization. It helps an agent turn PyTorch logic or an existing kernel into a high-performance GPU kernel through a structured, profile-driven workflow.

QLExpress is a powerful, lightweight, dynamic language for the Java platform aimed at improving developers’ productivity in different business scenes.