mobile-ai

Run local LLMs like Gemma, Qwen, and LLaMA on Android for offline, private, real-time chat and question answering with LiteRT and ONNX Runtime.

True on-device AI for Kotlin Multiplatform (Android, iOS, Desktop, JVM, WASM). LLM, Speech-to-Text and Image Generation — powered by llama.cpp, whisper.cpp and stable-diffusion.cpp.

On-device ComfyUI style node graph for Stable Diffusion on Android. Local SD1.5, SDXL, Flux and Z-image text to image, img2img, inpainting and video on the Qualcomm Hexagon NPU. Fully offline, no server, no account.

Pocket Lobster for Android: Codex, Claude Code, OpenMinis and a phone UI agent with multi-agent collaboration, three terminals, shared browser, Shizuku and accessible mobile workflows.

📱 Full AI development environment on Android. Ollama + Vector DB + API Server. One command. No root.

Kotlin-first framework for building Model Agnostic AI applications, RAG systems, and agentic workflows on Android and Kotlin Multiplatform.

Android local LLM runtime with GGUF, llama.cpp, OpenAI-compatible API, benchmarks, and device-aware recommendations.

Edge device object detection with Kotlin & TensorFlow Lite Mobile & edge AI object detection using Kotlin

HxSync: An Android AI Agent client with 4-tier scheduling (local/API/remote/gateway), MCP-compatible memory, and human-in-the-loop confirmations. Bring your own LLM.HxSync 是口袋里的 AI Agent 客户端：四档调度（本地/API/远端/网关），MCP 记忆内核，人在回路确认卡。自带模型，也可接你自己的 AI 服务。

Offline Android object detection app using LiteRT (Google AI Edge). Pick any photo from gallery → instant bounding boxes with labels & confidence scores. Built with Jetpack Compose + Clean Architecture.

Headless on-device LLM node for Android — runs a model on your phone's GPU via Google's LiteRT-LM and serves an OpenAI-compatible API (chat, embeddings, RAG, image-gen, TTS) over your LAN. Fork of Google's AI Edge Gallery, rebuilt as infrastructure you operate, not an app you open. AGPL-3.0.

VisionLite is a low-latency Android face vision toolkit for detection, alignment, embedding, anti-spoofing, and quality analysis.

A modular Android library for on-device LLM inference using GGUF models powered by llama.cpp. Load local language models, stream chat responses, and integrate offline AI into any Android app—no cloud, no API keys, and no data leaving the device. Built with Kotlin, Coroutines, and JNI.

The most powerful AI agent and AI chat software on Android/Operit是一款Android上能力最为强大、发展最久的AI Agent

Offline-first AI agent runtime and workspace for Android — Kotlin, Flutter, MCP, on-device LLM, and a policy-gated tool runtime.

Android dog-breed classification with Kotlin, Jetpack Compose, and TensorFlow Lite for on-device inference.

AI-Powered Flower Identification Android Application

Android + Flask mobile assistant demo powered by the OpenAI API.

Android dog breed classification app using Kotlin, Jetpack Compose, and TensorFlow Lite, with on-device inference and a simplified public demo UI.