ateymoori/oh-my-android
Native macOS control panel for the Android Emulator and real devices with a built-in stdio MCP server (18 tools) that lets Claude Code, Codex, Cursor and VS Code see and drive the emulator. MIT licensed.
Mobile-Artificial-Intelligence/maid
Free and open-source Android app for interfacing with llama.cpp models locally and remote APIs (Anthropic, DeepSeek, Mistral, Ollama, OpenAI). MIT licensed.
jev-chat/jev-chat-jarvis
Android reply copilot for WhatsApp that reads the chat on screen, judges intent and risk with your OpenRouter key, drafts two replies, and fills the one you pick. MIT licensed.
mobile-next/mobile-mcp
MCP server that lets AI agents automate and inspect iOS and Android simulators, emulators, and real devices through accessibility trees, screenshots, and structured device tools. Apache-2.0 licensed.
vadimchernets/poly-duo
iPhone Shortcut that chains the official Ask ChatGPT and Ask Claude app actions, so one model answers and the other reviews it, on the user's own accounts with no server or API keys; the Android counterpart is a separate app with its own key path and does not use the Shortcut's actions. Apache-2.0 licensed.
mlc-ai/mlc-llm
Deployment engine that compiles and runs LLMs across browsers, mobile devices, and local hardware.
k2-fsa/sherpa-onnx
Complete speech toolkit with ASR, TTS, diarization, source separation, and VAD across embedded and edge environments via ONNX Runtime.
google-ai-edge/LiteRT-LM
Production-ready runtime for deploying LLMs on edge devices with low-latency inference and optimized hardware paths for mobile and embedded platforms.
google-ai-edge/mediapipe
Cross-platform multimodal pipelines.
google-ai-edge/LiteRT
Google's production-ready on-device ML and GenAI deployment framework. Supports Android, iOS, Web, Desktop, and IoT targets with GPU/NPU acceleration via a unified edge-first runtime. Apache 2.0 licensed.
maziyarpanahi/openmed
Local-first healthcare AI framework and application for clinical text de-identification, entity extraction, and on-device specialized model serving. Apache 2.0 licensed.
moonshine-ai/moonshine
Open-source on-device voice AI toolkit for low-latency speech-to-text, intent recognition, and text-to-speech.
a-ghorbani/pocketpal-ai
Open-source app that brings small language models directly to your phone. Run AI 100% privately on iOS and Android with no cloud required. MIT licensed.
ChatGPTNextWeb/NextChat
Light and fast AI assistant supporting Web, iOS, macOS, Android, Linux, and Windows. One-click deploy with multi-model support. MIT licensed.
stablyai/orca
An agentic development environment (ADE) for running and orchestrating coding agents in parallel Git worktrees on desktop and mobile. MIT licensed.
tester-army/e2e
Agent-native web and mobile end-to-end testing framework that executes natural-language goals and records verified actions for replay without further model calls. Apache-2.0 licensed.
pingdotgg/t3code
Minimal web GUI and desktop app for interacting with coding agents like Codex, Claude Code, Cursor, and OpenCode. MIT licensed.
UbiquitousLearning/mllm
Fast and lightweight multimodal LLM inference engine for mobile and edge devices. Optimized for running vision-language models on resource-constrained hardware with efficient memory management. MIT licensed.
CherryHQ/cherry-studio
AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs. AGPL-3.0 licensed.
OpenBMB/MiniCPM-V
Compact vision-language model family with edge-focused deployment examples and strong OCR-oriented use cases.
elizaOS/eliza
Autonomous multi-agent framework for building and deploying AI-powered applications. Features Discord/Telegram/Farcaster connectors, RAG support, and a modern web dashboard.
alibaba/zvec
Lightweight, lightning-fast, in-process vector database from Alibaba. Built on Proxima (Alibaba's battle-tested vector search engine) for production-grade, low-latency similarity search. Apache 2.0 licensed.
genkit-ai/genkit
Open-source framework for building full-stack AI-powered applications in JavaScript, Go, and Python. Built and used in production by Google's Firebase. Unified interface for integrating AI models from multiple providers with built-in RAG, tool calling, structured outputs, and developer tools. Apache 2.0 licensed.
Browser-based photo upscaler and restorer: Real-ESRGAN upscaling up to 8x, GFPGAN face restoration, DDColor colorization and U2-Netp background removal run on-device through onnxruntime-web with a Rust/WebAssembly core, so photos are never uploaded. Works offline as a PWA and also ships as a browser extension and Mac/Windows/Android apps. MIT or Apache-2.0 licensed.