Pinned Loading

1. Kwipu Kwipu Public

   Ask questions across your Markdown notes using a fully local Graph RAG engine. Built for Obsidian vaults, works with any folder of Markdown files. Extracts entity-relation triples from wikilinks & …

   Python 589 60

2. writher writher Public

   🎙️ Offline voice productivity for Windows - dictate text into any app and control a local AI assistant by voice. Manages notes, to-do lists, appointments & reminders. 100% local: faster-whisper + O…

   Python 82 19

3. picchio picchio Public

   A streaming Mixture-of-Experts (MoE) inference engine for the GPT-OSS models (20B and 120B), written in pure C, that runs models larger than your RAM on ordinary consumer hardware.

   C 6 1

4. colizig colizig Public

   A from-scratch Zig inference engine for large MoE models - stream experts from SSD → RAM → VRAM and keep the resident working set small. ColiZig explores a different way to run Mixture-of-Experts m…

   Zig 3 1

5. TinySearch TinySearch Public

   Forked from TinySuiteHQ/TinySearch

   Shrink the web for your local LLMs!

   Python

6. retro-agent retro-agent Public

   A lightweight AI agent for legacy hardware. Written in Zig, runs on Windows XP SP3 x86 with 64 MB RAM. Connects to Ollama for LLM-powered system diagnostics and management via function calling.

   Zig 18 3