Your home as a live 3D model in the Mac menu bar and on iPhone — lights and ACs via Smart Life, Home Assistant, Philips Hue and Apple Home.
Headless control for Apple's Simulators — 3D models, taps, swipes, multi-finger gestures, 60 fps streaming, and a multi-device farm
Peekaboo is a macOS CLI & optional MCP server that enables AI agents to capture screenshots of applications, or the entire system, with optional visual question answering through local or remote AI models.
iPhone, iPad and CarPlay client for the Hermes AI agent or any OpenAI-compatible model. Talk or type, watch it work, approve what it runs. No account, no tracking.
Native iOS AI voice calls · Direct model APIs, MCP and your own backend · iOS 26+
Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cryptographic identity. Built in Swift. Fully offline. Open source.
Open-source on-device AI agent workspace designed for iOS/iPadOS and Android: Linux environments, code execution, local web previews and supported local models. Notes, Office, creative tools and BYOK cloud AI. iPad-first. 开源设备端 AI 智能体与本地运行环境。
The free AI already on your Mac. CLI tool, OpenAI-compatible server, and interactive chat — all on-device via Apple Intelligence. No API keys, no cloud, no downloads.
Small enough to run on your Mac. Smart enough to ask. An open-source computer use agent for macOS, with local models (MLX) that see, decide and act. 得心,应手。
Local AI, native to your Mac. Chat, serve, monitor, and connect MLX models from one macOS app.
A native macOS harness for coding agents that works with any model, on the opencode engine.
An API-compatible, drop-in replacement for Apple's Foundation Models framework with support for custom language model providers.
Remove unwanted Apple Intelligence models without disabling SIP. CLI, nix-darwin, and Home Manager support.
Run a 105 GB AI model on a Mac that can't hold it. Slotstream streams Qwen3.8-Flash-Next (125B mixture of experts) from your SSD and caches the busiest experts in memory, so it runs on Macs with 16 to 64 GB. One native Swift binary on MLX and Metal, no Python, offline. Works with Claude Code, Codex and Ollama or OpenAI clients.
This is a personal repo for porting new Photographic Styles in iPhone to older model below 15
The official Swift SDK for Model Context Protocol servers and clients. Maintained in collaboration with Apple
Community edition of RepoPrompt: a native macOS context engineering app for AI coding agents, with an MCP CLI.
Emerging and experimental patterns for building with the Foundation Models framework
computer-use assistant w/ decision models
⚡ Native MLX Swift LLM inference server for Apple Silicon. OpenAI-compatible API, SSD streaming for 100B+ MoE models, TurboQuant KV cache compression, MACOS + iOS iPhone app.
Swiftlet is a Swift and Metal runtime that runs large Qwen Mixture-of-Experts models locally on Apple devices by streaming expert weights from storage, enabling 35B and 80B models to run with low RAM, including on iPhone.
Give AI agents their own macOS displays. Keep your screen, cursor, and focus.
Native macOS video editor that coding agents can drive: everything the UI does, Claude Code and Codex can do through the bashcut CLI or MCP server. Layered timeline, captions, voiceover, plugins.
Type-safe tools, on-device inference, multi-agent workflows, memory, and guardrails, in one Swift-native runtime
Local computer use on Apple silicon: on-device typed decisions with laya and form filling with CUA-S1-FORMS, driven through the macOS Accessibility API
Notes for iPhone and Mac that ChatGPT, Claude, Claude Code and Codex can read and edit. Feels like Apple Notes, imports it too. Open source, encrypted.
Notational Velocity: modeless, mouseless Mac OS X note-taking application
Core ML model zoo for iOS/macOS — PyTorch models converted to ready-to-use .mlpackage, each with a conversion script and SwiftUI sample app. Sibling repos cover Apple's Core AI framework (iOS/macOS 27) and on-device LLMs.
Say it, and your Mac does it. A computer-use harness on Jev that reads the screen through Accessibility. Fast, no vision model
A lightweight, native Swift 6 bridge integrating TypeSafe AI's Jev System One decision model into Apple's Foundation Models framework.
Your feedback layer for AI collaboration. Point at anything on your Mac - text, screenshots, web elements, or voice - and your agent reads and resolves your comments over MCP.
Annotate a running iOS simulator in your browser, and hand the notes to your coding agent.
The local AI app built for the Mac. Run LLMs like Qwen, Gemma and GPT-OSS on Apple Silicon, including models far bigger than your Mac's memory, by streaming MoE experts from the SSD.
Run Desert Ant Labs on-device models from the terminal.
A native macOS lab for teaching tiny language models to think — build the architecture, train the weights, and watch a small LLM emerge from scratch, locally on Apple Silicon with custom data, tokenizers, checkpoints, and MLX acceleration.
670 MB/s Swift tokenizer inspired by Marcel Roed’s gigatoken. Zero-copy, NEON/CRC32, WebAssembly, and Embedded Swift.
Template iOS app using Clean Architecture and MVVM. Includes DIContainer, FlowCoordinator, DTO, Response Caching and one of the views in SwiftUI
Peer-to-peer cluster implementation for Swift Distributed Actors
Experimental macOS 27 App Intents model delegation provider backed by Claude Code
An experimental iPhone companion that gives voice models a face and personality, built with SwiftUI, WebGL, WebRTC and Blender.
A visionOS app to preview large scale models in real life (like housing plans!)
Runs openpilot's large driving model on Jetson Orin Nano, macOS, iOS, Android, CUDA laptops with only your Comma, and a USB3 cable
Drive a real browser from the command line and from MCP. Stable element refs, zero dependencies.
Native macOS spatial slideshows using installed Apple Photos Reframe models, with optional local edge expansion.
Grux AI: open source AI agent for macOS (MIT). One Command Panel for your mail, calendar, meetings, files and a shell it can undo. Runs on a local Ollama model with no key, or your own key. Change it with Claude Code. Download Grux.dmg at gruxai.com, or brew install --cask dotcomjack/tap/grux
Fast, private, on-device dictation for Apple silicon Macs
works well with rocks; native Apple Silicon inference engine for capable and tiny models (ternary, Qwen) via MLX Swift. Hadamard-rotated KV cache, vision, tool calling, OpenAI + Anthropic APIs
📐 Describe a part, get a printable model: a native macOS CAD studio powered by Claude Code and build123d.
Example macOS app and CLI to generate short clips, fully on device.
SmartCodable is a data parsing library built on Swift’s Codable, designed for simple usage and strong real-world compatibility. It gracefully handles missing fields, default values, and evolving JSON structures. SmartCodable 是基于 Swift Codable 的数据解析库,主打简单易用与真实业务场景下的强兼容性,能够优雅应对不断变化的 JSON 数据。
MLXUI is a Local AI Browser for MLX. Run MLX models AI Workflows.
Open-source, bring-your-own-key (BYOK) multi-model AI chat client for iOS, Android and the web. One LLM client for OpenAI, Anthropic, Google Gemini, OpenRouter, DeepSeek and ten more providers, plus any OpenAI-, Anthropic- or Gemini-compatible endpoint — Ollama, LM Studio, llama.cpp, vLLM. Local-first, self-hostable, no account. AGPL-3.0.