Trending AI projects

Apache TVM is an open machine learning compilation framework that optimizes and deploys models into minimum deployable modules using a Python-first approach.

UniFi MCP provides MCP servers and plugins that let AI agents query and control Ubiquiti UniFi network, camera, and access systems via natural language, with safety, lazy tool loading, and remote relay options.

Relaticle is an open‑source, self‑hosted CRM built with Laravel that exposes a REST API and a custom MCP server so AI agents (Claude, ChatGPT, etc.) can read and write CRM data. It offers customizable fields, workspace isolation, per‑field encryption, and a modern UI, targeting AI‑forward teams that want full data control.

NetCoreKevin is an enterprise‑grade AI‑agent platform built with .NET 9 and Vue 3, offering multi‑tenant RAG, dynamic skill scripts, memory, token quotas and OpenAI‑compatible APIs, plus a full admin UI and many pluggable infrastructure modules.

A notebook-based tutorial that takes learners from PyTorch implementations of LLM algorithms down to Triton kernels and CUDA/system optimization, with multiple learning routes and hands-on experiments for training, inference, and performance measurement.

Tingly Box is a high-performance AI gateway that unifies disparate agent and model protocols, providing teams with centralized management, usage tracking, and remote control capabilities.

A curated GitHub list of generative AI projects and services, organized by category (text, coding, image, audio, etc.), with recommended reading, milestone announcements, and community-contributed links.

An immersive AI companionship assistant that combines LLMs with emotion recognition, desktop vision, and Galgame-style visuals to create a reactive digital character.

A code-free visual web crawler and browser automation tool that allows users to design and execute data collection tasks through a point-and-click interface.

A Pi extension that enables AI agents to interact with desktop applications by observing UI elements and performing actions like clicking and typing across macOS, Windows, and Linux.

An AI-powered penetration testing agent platform that autonomously scans, exploits, and reports on authorized targets using LLMs (OpenAI/Anthropic), with an attack-graph-driven self-improving loop, a web console, and strong built-in safety restrictions.

An open-source PDF translation tool that preserves original layout and formulas, specifically optimized for scanned and image-based PDFs.

A collection of interactive ComfyUI nodes that add 3D Gaussian generation, infinite-canvas image editing, and 3D pose studio capabilities directly into the node-based workflow.

Jeff is a 0.8 B “System 1” model that makes ultra‑fast, calibrated classification decisions. It uses lightweight LoRA adapters for specific domains (guard, triage, spam, etc.) and can act as a fast‑path filter in front of a large LLM (e.g., Qwen 3.8‑27B), handling most queries in ~30 ms and delegating only uncertain cases. Weights and adapters are open on Hugging Face, GGUF files enable llama.cpp use, and a simple HTTP/Python API (`jeff‑serve`) lets you plug Jeff into any service.

A local server that enables Claude Code to run on Apple Silicon Macs using open-source models via the MLX framework, eliminating the need for cloud API keys and ensuring data privacy.

An AI-powered tool that extracts individual garments from photos to create a digital wardrobe and generates modeled editorial previews of outfits.

Orchard is a native macOS application that provides a visual dashboard for managing Apple's container runtime, including support for local AI model integration, Kubernetes clusters, and Linux virtual machines.

A suite of AI agent skills for calculating and analyzing Vedic astrology charts directly from birth details, featuring specialized modules for career, love, and birth time rectification.

A development environment and orchestration framework for building, testing, and deploying collaborative multi-agent AI systems using declarative configurations.

OpenSpline is a developer toolkit that adds synchronized, talking-head video avatars to existing voice AI agents, supporting both cloud and self-hosted GPU deployments.

A curated, evidence-based curriculum and benchmark suite for understanding modern CPU performance, complete with an MCP server for AI-assisted performance engineering.

MobileBuildMCP is an open‑source tool that runs an MCP server and a CLI to let AI agents (or humans) invoke Xcode build and test commands for iOS/macOS projects. Install via Homebrew or npm, start the server with `mobilebuildmcp mcp`, and use subcommands like `simulator build` to compile or test apps. Optional “skills” JSON files teach LLM assistants how to talk to the server. MIT‑licensed, macOS‑only, with Sentry telemetry.

An AI-powered tool that generates complete multi-page social media posts, including outlines, copywriting, and visually consistent images, from a single sentence.

The official Python client and CLI for the Hugging Face Hub, enabling users and AI agents to programmatically download, upload, and manage ML models, datasets, and spaces.

StoryForge is an open‑source, local‑first web app for AI‑assisted writing and adaptation of novels, scripts, comics, motion‑drama packs and world‑building, with a unified React/TS UI, IndexedDB storage, a central Harness for versioned AI task logging, plugin support and optional cloud or local LLM integration.

OpenArm is an open‑source 7‑DOF humanoid robotic arm (≈$6.5k) with a standardized test‑cell, safety‑focused hardware, and a full software stack (ROS‑2, dora‑rs, MuJoCo, Isaac Lab). It provides CAD, robot description, CAN drivers, teleoperation, dataset tools, and simulation models, enabling reproducible physical‑AI research and easy switching between real and virtual arms.

suyuan-skill is a collection of three practical skills for AI agents: whatis turns confusing concepts into simple, evidence-backed diagrams; break-ai-slop forces the agent to think like an expert before answering open-ended tasks; get-biji migrates your Get Notes (biji.com) recordings and notes to local Markdown files and uses Get's transcription quota for audio/video.

libcom is a Python toolbox that bundles state‑of‑the‑art models for every step of realistic object insertion—placement scoring, colour/illumination harmonisation, shadow & reflection synthesis, and even generative composition—so you can go from a raw foreground‑background pair to a polished composite with a few function calls.

An AI-powered software engineering platform that uses multi-agent systems and knowledge graphs to automate bug fixing, feature implementation, and codebase navigation.

An agentic workspace that integrates AI agents side-by-side with real applications and chat tools, featuring local-first memory and shared tool access.

An Android application that turns a mobile phone into a local, OpenAI-compatible LLM server using the LiteRT-LM runtime.

An open-source agentic BI platform that uses a governed context layer to power dashboards, AI agents, and data apps with a software-engineering workflow.

A high-performance Rust-based web research tool for AI agents that provides stealthy fetching, keyless search, and crawling without requiring API keys.

A lightweight system that helps developers align AI agents with their codebase's specific standards and patterns to ensure consistent, high-quality builds.

An agentic AI development platform that helps developers plan, build, and review software by turning prompts into executable specs and using parallel agents across multiple interfaces.

A local-first AI writing workbench for creators that integrates specialized agents and a diff-based review system to collaboratively draft and refine manuscripts.

Paper2Patent is an open‑source AI‑assisted pipeline that converts a research paper into a full Chinese invention‑patent draft (five sections, claims, drawings, and compliance checks). It provides prompt templates, a reusable “Skill” for Codex/Claude/Cursor/Windsurf, and Python scripts for auto‑checking, drawing generation, and DOCX/PDF export.

A virtual pixel-art office where local LLM-powered agents autonomously collaborate, hire new team members, and execute tasks with persistent semantic memory.

A library of skills and plugins that enables AI agents to automate Salesforce development tasks, including generating code, managing objects, and deploying applications.

A project management framework and continuity engine for fiction writers that uses AI agents to draft, track, and validate narrative consistency in long-form stories.

dshfind is a community site for DeepSeek Harness (DSH) that offers structured lessons on DSH principles, a plugin marketplace that automatically aggregates GitHub repositories tagged `dsh-plugin`, and best-practice guides with community rankings. It solves discovery and education gaps for the DSH ecosystem.

NovelToGame is an open‑source set of AI‑agent skills that turn a full novel into a playable, source‑grounded game (web‑browser or 3D) with full design documentation and a single‑run QA. It works with Claude Code, Codex, or Kimi Code, requires no bundled engine, and records every mechanic’s citation to the original text.

A local desktop workspace that integrates multiple AI coding agents like Claude Code and Grok into a single interface with built-in design tools.

Zvec is a lightweight, high-performance in-process vector database designed for embedding directly into applications to enable fast similarity and hybrid search without external server infrastructure.

A plugin marketplace and discovery platform for Claude Code that provides curated agents, commands, and hooks to extend AI development workflows.

Zotero‑AI‑Butler is an open‑source Zotero plug‑in that lets you call LLMs (Gemini, OpenAI, Claude, etc.) to automatically read PDFs, generate Markdown summaries, mind‑maps, one‑page visual abstracts, and batch literature‑review drafts. It works via a task queue, supports multi‑round prompting, Base64 PDF upload for multimodal models, and offers a sidebar preview with LaTeX rendering. Users supply their own API keys; the plug‑in stores no data. Installation is a simple .xpi download, and the UI provides right‑click shortcuts, a dashboard, and customizable prompt templates.

A collection of production-oriented workflow plugins for Claude Code that automate the creation of projects, assets, and professional documents.

An open-source library for intelligent LLM routing that dynamically selects the most cost-effective and performant model for each query using over 16 routing strategies.