Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust.
Safe rust wrapper around CUDA toolkit
Unified Rust framework for building, verifying, and testing web apps. Kernel-based architecture for Lean 4 verification and GPU acceleration. Built-in red-teaming tools for automated vulnerability detection.
Pure Rust + CUDA LLM inference engine — no PyTorch, OpenAI-compatible, serves Qwen3 to Kimi-K2