A self-hosted AI agent platform that turns spare hardware into a distributed compute cluster, providing a framework-agnostic memory system and a full web-based desktop interface.

A Python speech-to-text library that provides fast, real-time transcription with integrated voice activity detection and wake word support.

A node-based visual programming framework for 3D reconstruction and computer vision, enabling users to build complex data processing pipelines using AI and photogrammetry.

A framework for building federated AI systems that enables decentralized model training across various ML frameworks and devices.

An MCP server that gives AI agents real-time web access with built-in bot and CAPTCHA bypassing, offering 69 tools for search, scraping, and structured data extraction.

Neovim plugin that connects to an OpenCode AI server, injecting editor context (buffer, selection, diagnostics, etc.) and exposing Ask/Select/Prompt/Operator commands. Provides built‑in prompts (explain, fix, test, …), real‑time edit diffs, permission handling, and event autocmds. Install via vim.pack, lazy.nvim, or nixvim; configure server URL or auto‑start command; use example keymaps like `<C‑a>` to ask the AI.

An AI-driven cultivation world simulator where every NPC is an independent LLM agent, creating an emergent narrative based on a rigorous set of cultivation rules.

A full-stack development platform for AIoT and HMI products that decouples hardware from software to enable scalable app deployment and AI agent integration on ESP devices.

DEIMv2 is a real-time object detection framework leveraging DINOv3 features to provide scalable, state-of-the-art performance across various model sizes for edge and server deployment.

A multi-agent RAG customer support system built with LangGraph and LangChain that features security guardrails and a human-in-the-loop approval process for sensitive operations.

A GPU-native SQL engine that accelerates databases like DuckDB by offloading query execution to NVIDIA GPUs without requiring query rewrites.

An AI-powered internationalization framework for web and mobile apps that provides per-component content management, a visual editor, and a CMS.

Docspell is an open‑source personal document management system that organizes scanned papers, emails and other files. It uses Stanford CoreNLP to automatically suggest tags, dates and correspondents, runs OCR when needed, and provides full‑text search via a REST API and a mobile‑friendly Elm web UI. Installable via Docker, Debian packages, Nix or Helm, it is licensed under AGPL‑v3.

A performance benchmarking tool for NVIDIA GPUs that measures memory bandwidth and latency across CPU-to-GPU and GPU-to-GPU links, including multi-node configurations.

An attention-free RNN architecture that delivers Transformer-level LLM performance with linear-time complexity and constant VRAM usage.

An open-source, local-first AI journal that uses a multi-agent system to automatically organize life fragments into a structured, private knowledge base.

A local knowledge base QA system that uses two-stage retrieval to provide accurate answers from various file formats while supporting fully offline deployment.

A collection of distilled LoRA checkpoints for MiniMax-H3 that enables fast, few-step video and audio generation via Diffusers and ComfyUI.

A token-efficient MCP server that uses tree-sitter AST parsing to allow AI agents to retrieve precise source code symbols rather than reading entire files.

A Docker CLI plugin that allows users to build, run, and share AI agents using declarative YAML configurations without writing code.

A real-world reinforcement learning framework for robotic manipulation that enables a unified pipeline from imitation learning to offline/online RL and one-step policy distillation.

A scientific machine learning library for solving differential equations and learning operators using physics-informed neural networks and other advanced AI architectures.

A desktop workspace and GUI for OpenCode, providing a professional interface with multi-instance sessions, voice input, and integrated local web tools.

ArtCraft is a desktop IDE that lets artists build 2‑D or 3‑D scenes and then fill them with AI‑generated images, video, audio, meshes and worlds. It offers visual tools for compositing, pose transfer, background removal, and prompt‑driven generation, supporting a catalog of 62 built‑in models plus external providers like Midjourney and Sora. Available for Windows/macOS (Linux from source), it targets creators who want repeatable, controllable AI art workflows.

A Doom-based AI research platform for developing and testing bots that play the game using only visual information, primarily for reinforcement learning research.

A Model Context Protocol (MCP) server that connects LLMs to ROS-based robots, enabling natural language control and real-time monitoring without modifying robot source code.

A Python package for audio stem separation that uses UVR models to split audio files into vocals, instrumentals, and other specific instrument tracks.

A terminal-native agentic development environment that uses a blackboard model to coordinate multiple AI coding agents across a shared kanban board and dependency graph.

Akagi is a Rust‑based desktop assistant for online riichi mahjong (Mahjong Soul, Tenhou, etc.). It captures game traffic via a MITM proxy or Chromium DevTools, runs a built‑in neural‑net bot (and optional cloud‑inference model) to compute shanten, waits, win probabilities, opponent risk, and suggested discards, and shows these numbers in a draggable HUD. Completed matches are stored locally and visualised in a History tab with rank pie charts, PT trend lines, and detailed stats. The app is distributed as a single portable binary for Windows, macOS (Apple Silicon) and Linux, uses Tauri 2 + React for the UI, and is licensed under Apache 2.0.

An open-source evaluation framework for AI applications and agents that provides a library of ready-to-use graders and tools to generate custom evaluation rubrics.

An open-source WordPress plugin and MCP server that gives AI agents direct access to execute PHP, run WP-CLI commands, and manage files on a WordPress site.

Reckless is a high-performance, open-source competitive chess engine that ranks among the top engines in major tournaments like CCC and TCEC.

An MCP server that integrates the Camoufox anti-fingerprinting browser with AI assistants to perform advanced JavaScript reverse engineering and JSVMP analysis.

An AI companion mod for Minecraft that executes complex tasks through natural language, acting as a server-side player capable of mining, building, and fighting.

A lightweight, self-contained Rust-based coding agent for the terminal that supports both cloud LLMs and native offline local inference.

A high-performance, Rust-powered data engine designed to process multimodal AI workloads, allowing users to handle images, audio, and video alongside structured data at scale.

A tool that lets AI agents control your actual browser and macOS apps using your existing sessions and logins, avoiding the detection and setup hurdles of traditional automation.

A unified proxy server for AI CLI tools that aggregates multiple AI subscriptions and providers into a single managed API layer with routing and quota control.

A privacy-first browser tool that exports Claude.ai conversations to high-fidelity Markdown, optimized for Obsidian and RAG pipelines.

A comprehensive toolkit for large model compression that provides unified access to quantization, speculative decoding, and sparse attention to reduce model size and increase inference speed.

Olares is an open-source personal cloud operating system that enables users to run AI agents and LLMs locally on their own hardware using Kubernetes.

A privacy-first Android application wrapper for the Lumo AI assistant that adds native features like speech-to-text and in-app billing to the web experience.

Humanity's Last Exam (HLE) is a multi-modal benchmark consisting of 2,500 expert-developed questions across various academic subjects to test the limits of AI knowledge.

A multi-model AI agent desktop client that unifies 17+ AI providers into one interface with persistent memory, remote control via messaging apps, and MCP support.

An open-source voice dictation tool that converts speech to text anywhere on your system, supporting local transcription and LLM post-processing for privacy and speed.

A unified simulation infrastructure that integrates CARLA and AirSim into a single process to enable high-fidelity, synchronized air-ground embodied intelligence research.

Vespa is an open‑source platform for large‑scale search and AI serving, offering vector/tensor indexing, real‑time model inference, and low‑latency query processing. It can be run on Vespa Cloud or self‑hosted, with extensive docs, sample apps, and a permissive Apache 2.0 license.

KubeSphere is a distributed operating system for cloud-native application management that simplifies Kubernetes management across multi-cloud, edge, and datacenter environments.