NobodyWho Chat：完全离线的本地 AI 助手

NobodyWho Chat is a fully offline, private AI assistant that runs entirely on your device. Every conversation happens locally: no data leaves your phone, no internet connection is required, and there are no accounts, subscriptions, or hidden costs.

Built with NobodyWho, a library designed to run LLMs locally and efficiently on any device, using the wonderful llama.cpp under the hood.

Features

Fully Private & Offline: all conversations and data stay on your device; no conversation is collected, shared, or sent to the cloud, and the chat works without an internet connection
Free & Open: no sign-ups, subscriptions, or paywalls
Chat — stream responses from a local LLM in real time
Industry-leading models — download and use several powerful LLMs, including thinking models
Tool calling: give the model access to custom functions (e.g. weather, historical facts, stock prices...)
Vision & Hearing — image & audio ingestion with a multimodal model

1. Getting Started

Install dependencies with npm install.

For iOS, install pods: cd ios && pod install && cd ..

2. Sentry

Inside android and ios folder, create sentry.properties file and paste the following:

defaults.url=https://sentry.io/
defaults.org=example-org
defaults.project=example-project
auth.token=sntrys_YOUR_TOKEN_HERE

3. Run the App

# Android
npm run android

# iOS
npm run iOS

Note:

For iOS, if you have issues with metro, run npm start and then run the project on Xcode.

The script Copy Optional Assets in Xcode > Build Phases copies files in assets folder (in debug mode only). On Android, this step is achieved with assets.srcDirs += ["$rootDir/../assets"] in build.gradle.

Miscellaneous

Update tests

npm run test-update

iOS cleanup

make ios-clean

Watchman/Metro/Cache cleanup

make clean-metro
make clean-watchman
make clean-cache

4. Build the App

Build Android

make android-apk

Install apk on android

adb install android/app/build/outputs/apk/release/app-release.apk

make android-aab

For iOS, run prod scheme on Xcode.