Local Chat: Offline AI
📱 Overview
Local Chat runs large language models 100% on the phone with llama.cpp. Download a model once, then chat without internet. Nothing you type ever leaves the device.
✨ Key Features
On-device inference
llama.cpp with Metal GPU offload on iPhone, real token streaming, instant Stop and KV-cache reuse between turns.
35 curated models
Everyday, Study, Business, Coding, Vision and more, with a "Fits my phone" filter and Hugging Face GGUF search.
Vision & documents
Ask about photos with vision models; attach PDF, TXT, CSV, JSON or code, with text extracted on the device.
Reasoning models
Collapsible "think" panel for Qwen3, DeepSeek R1 and others, plus assistant modes and resumable downloads.
🛠️ Technical Stack
-
👨💻 Project Details
🔹 Category: On-device AI / LLM
🔹 Package: com.local.chatai
🔹 Publisher: Balliram App Development
🔹 Role: Design, development and Play Store release
🔹 Status: Live on Google Play