Local LLMs that run lean, load instantly and keep getting smarter. A from-scratch runtime + model format in Zig: zero-copy mmap, CPU/CUDA/Vulkan engines, OpenAI-compatible server, 2:4 sparsity, Net2Net growth, in-file version history, self-learning loops.
sparsity ai vulkan zig cuda inference mmap quantization model-format self-learning net2net huggingface llm safetensors gguf openai-compatible model-container
-
Updated
Oct 22, 2026 - Zig