Nexa SDK is a comprehensive toolkit for supporting GGUF and MLX model formats. It supports LLM and VLMs

An AI-powered file management tool that ensures privacy by organizing local texts, images. Using Llama3.2 3B and Llava v1.6 models with the Nexa SDK, it intuitively scans, restructures, and organizes files for quick, seamless access and easy retrieval.

[ICLR 2019] ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware

Official PyTorch implementation of "EdgeSAM: Prompt-In-the-Loop Distillation for On-Device Deployment of SAM"

TinyChatEngine: On-Device LLM Inference Library

Android on_device 1:1 Face Recognition 、 Liveness detection and 1:N & M:N Face Search SDK 🧒 离线版设备端Android1:1人脸识别 动作活体、静默活体、近红外双目活体检测 以及1:N M:N 人脸搜索算法SDK 封装。全程可开飞行模式不用联网

[CVPR 2025] Official PyTorch implementation of "EdgeTAM: On-Device Track Anything Model"

On-device Neural Engine

On-device LLM execution in React Native with Vercel AI SDK compatibility

[ECCV 2018] AMC: AutoML for Model Compression and Acceleration on Mobile Devices

Android Input Method Editor (IME) based on Whisper

NativeMind: Your fully private, open-source, on-device AI assistant

Android Voice Activity Detection (VAD) library. Supports WebRTC VAD GMM, Silero VAD DNN, Yamnet VAD DNN models.

A ready-to-use, minimal app that converts any speech into text.

AubAI brings you on-device gen-AI capabilities, including offline text generation and more, directly within your app.

The world’s #1 end-to-end job agent: semantic filters, ATS-optimized resumes, referrals from hiring managers — 100% hands-free.

Multi-Task Temporal Shift Attention Networks for On-Device Contactless Vitals Measurement (NeurIPS 2020)

电子鹦鹉 / Toy Language Model

[ECCV 2018] AMC: AutoML for Model Compression and Acceleration on Mobile Devices

Source code for my blog post series "On-device training with Core ML"