RyanCodrai / turbovec

A vector index built on TurboQuant, written in Rust with Python bindings

warpfront / hipfire

RDNA-native LLM inference engine in Rust.

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.

Bellman281 / esp32-tinyLLM

Rust inference engine for a 28.9M-parameter PLE TinyLM on the ESP32-S3 — 9.4 tok/s, faster than the C implementation it was ported from