Run frontier MoE models on hardware you already own — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

26 different problems dev teams faces with Git

A Straightforward, Step-by-Step Implementation of a Video Diffusion Model

Implementation of all RL algorithms in a simpler way

Building an advanced Agentic eCommerce WhatsApp bot

Agentic RAG to achieve human like reasoning

A Deep Thinking RAG Pipeline to Solve Complex Queries

Self improving agentic rag pipeline

Training architecture for self-improving AI agents.

Core concepts - where to apply parallelism in agentic solution

Deep research agentic system using Time Test Diffusion

A scalable RAG platform combining LangGraph agents, hybrid retrieval (Vector+Graph), and Ray orchestration on Kubernetes.

Building a Tiny GPU to understand AI hardware engineering

A multi-agent system trained with GRPO for reliable long-horizon task planning and execution.

Building a Senior Staff Engineer with Sub-Agent Teams in Claude Code

Building a Production-Grade MCP Server Architecture with a Multi-Agent System

A step by step implementation to build an agent that runs for week to complete a task and self improve.

Handling 10M+ docs using RAG with zero hallucinatons

An autonomous agentic pipeline that finds, proves, and patches real C memory-safety vulnerabilities end-to-end using a single 7B open-source LLM (Qwen2.5-Coder) on vLLM, with AddressSanitizer as a …

35 production-grade agentic AI architectures (Reflexion, LATS, GraphRAG, MemGPT, Voyager, BrowserAgent, ...) — a Python library and runnable textbook with multi-provider LLM support and a 17-task b…

A straightforward method for training your LLM, from downloading data to generating text.

From-scratch reimplementation of Google TimesFM (time-series foundation model) plus the full pretraining pipeline Google never open-sourced. Trained and honestly evaluated at 70M.

GLM-5.2, a 744 billion parameter mixture of experts model, in a pure C inference engine: quantized to int4, experts streamed from disk, deployed and benchmarked. Generates in 16 GB of RAM.

A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.

Vibe-Coding is easy. DevOps is hard. OpenCode & Git token-efficient AI agent automation for your app, business, and personal development. Opinionated tools, services, CLI & API stack for speed, sec…

This revolutionary software enables EV battery packs to be easily reused for stationary storage in combination with solar inverters

Nextcloud MCP Server

Local-first healthcare AI: clinical NER & HIPAA PII de-identification that runs 100% on-device. 2,200+ medical models, 21 languages, Apple MLX + Python, no cloud, no patient data leaving your netwo…

The open-source AI voice studio. Clone, dictate, create.