ODS v2.5.3 Bundles llama-server, Agents and n8n Into a One-Command Private AI Server — With AMD Strix Halo Unified Memory Support
GitHub·low signal
ODS turns a PC, Mac or Linux box into a private AI server in a single install, bundling llama-server inference, Open WebUI, a GPU monitoring dashboard, speech-to-text and text-to-speech, RAG and document search, image generation, and autonomous agents. Its hardware auto-detection spans NVIDIA, AMD including Strix Halo unified memory, Apple Silicon via Metal, and Intel Arc, selecting models to match. It defaults to Hermes Agent for agents with n8n supplying 400+ workflow integrations; Apache-2.0 at 3,764 stars and 543 forks.