Native macOS and Apple Silicon inference server exposing OpenAI, Anthropic and Ollama APIs, with a Swift app, written in Zig.
Serves language models from a Zig binary on Apple Silicon and NVIDIA GPUs using Metal and CUDA kernels.
Native macOS and Apple Silicon inference server exposing OpenAI, Anthropic and Ollama APIs, with a Swift app, written in Zig.
Serves language models from a Zig binary on Apple Silicon and NVIDIA GPUs using Metal and CUDA kernels.