Search Packages

smg-grpc-proto

Engine-agnostic LLM gateway in Rust. Full OpenAI & Anthropic API compatibility across SGLang, vLLM, TRT-LLM, OpenAI, Gemini & more. Industry-first gRPC pipeline, KV cache-aware routing, chat history, tokenization caching, Responses API, embeddings, WASM plugins, MCP, and multi-tenant auth.

smg-grpc-servicer

Engine-agnostic LLM gateway in Rust. Full OpenAI & Anthropic API compatibility across SGLang, vLLM, TRT-LLM, OpenAI, Gemini & more. Industry-first gRPC pipeline, KV cache-aware routing, chat history, tokenization caching, Responses API, embeddings, WASM plugins, MCP, and multi-tenant auth.

mooncake-transfer-engine

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

auto-round

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

gptqmodel

LLM model quantization (compression) toolkit with HW acceleration support for Nvidia, AMD, Intel GPU and Intel/AMD/Apple CPU via HF, vLLM, and SGLang.

auto-round-nightly

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

auto-round-lib

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

mooncake-transfer-engine-cuda13

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

dgxarley

Ansible playbooks for a 3-node K3s cluster with NVIDIA DGX Spark nodes for distributed LLM inference

terradev-cli

Cross-Cloud Compute Optimization Platform with Migration & Evaluation - v4.0.12

strands-sglang

SGLang model provider for Strands Agents for on-policy agentic RL training.

smg

Engine-agnostic LLM gateway in Rust. Full OpenAI & Anthropic API compatibility across SGLang, vLLM, TRT-LLM, OpenAI, Gemini & more. Industry-first gRPC pipeline, KV cache-aware routing, chat history, tokenization caching, Responses API, embeddings, WASM plugins, MCP, and multi-tenant auth.

mooncake-transfer-engine-non-cuda

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

infergrid

Tenant-fair LLM inference orchestration on a single GPU. No Kubernetes.

auto-round-hpu

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

agentic-coding-bench

Open-source benchmark for LLM inference on agentic coding workloads

llm-cal

LLM inference hardware calculator — architecture-aware (MLA/NSA/MoE), engine-aware (vLLM/SGLang), honest-labeled. Reads real safetensors bytes; supports 53 GPUs (NVIDIA / AMD / Huawei Ascend / 沐曦 / 昆仑芯 / 壁仞 / 寒武纪 / 海光).

kvcached

Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond

kvwarden

Tenant-fair LLM inference orchestration on a single GPU. No Kubernetes.

auto-round-kernel

A SOTA quantization algorithm for high-accuracy low-bit LLM inference, seamlessly optimized for CPU/XPU/CUDA, with multi-datatype support and full compatibility with vLLM, SGLang, and Transformers.

anchor-vision

Python client for Anchor — PaliGemma2 multi-LoRA vision inference

docvision

Production-ready document parsing with Vision Language Models

flashtts

基于SparkTTS、OrpheusTTS等模型，提供高质量中文语音合成与声音克隆服务。

terradev-mcp

Complete Agentic GPU Infrastructure for Claude Code