magnitudedev / magnitude

Open source inference engine that tunes its kernels on your hardware to find the fastest way to run each model. 2x faster than Ollama. Works on Apple Silicon, NVIDIA, AMD, or just a CPU.

midagedev / bloomery

Hybrid GPU + CPU inference for mixture-of-experts models, written in Rust down to the CUDA kernels. A drop-in llama-server.

h5i-dev / h5i

An Axum-based Rust web framework for building and verifying web apps with Lean 4, featuring optional GPU acceleration and integrated development infrastructure for red-teaming and CI/CD.

pegainfer-project / pegainfer

Pure Rust + CUDA LLM inference engine — no PyTorch, OpenAI-compatible, serves Qwen3 to Kimi-K2

Dr-TSNG / altdb

A KernelSU module that provides alternative adb server over local network

VeryBaaad / ZNNext

A from-scratch, standalone implementation of the Zygisk Next module

Rust-GPU / rust-cuda

Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust.

hermit-os / kernel

A Rust-based, lightweight unikernel.

chelsea0x3b / cudarc

Safe rust wrapper around CUDA toolkit

antonellof / frink

Pure-Rust GGUF inference engine with quantized CPU, Metal & CUDA kernels, MoE support, OpenAI-compatible server. Benchmarked head-to-head against llama.cpp.

tsirysndr / bsdkrun

Instant MicroVMs/Unikernels for macOS and Linux