Similar apps
Compare allApp info
- Installs
- 10+
- Category
- Communication
- Content rating
- Everyone
- Version
- 2.2
- Released
- Aug 24, 2025
- Last updated
- 8 Apr 2026
- In-app purchases
- No
- Contains ads
- No
- Promo video
- No
- Requires Android
- 8.0
Loading…
This app isn't in any of our charts yet.
Release 2.2: -Corrected crashing if trying to load an unsupported model. -Updated llama.cpp Release 2.1: -Updated llama.cpp Release 2.0: -Added compatibility to dark theme. -Fixed empty saved conversations. Release 1.9: -Corrected thinking box going away after thinking. Release 1.8: -Added a delete all Conversations button. -Corrected duplicate of conversation when regenerating. Release 1.7: -Added a Thinking Box expandable. -Corrected history. Release 1.6: -Corrected main layout.
Chat with any light GGUF model from "HuggingFace" Offline
Offline AI Chat: Your Private AI Assistant Talk to an advanced AI chatbot anytime, anywhere completely offline. No internet connection is required after the initial model download, no data is ever sent to servers, and zero ad tracking. Offline AI Chat runs directly on your device, giving you full privacy and control. Key Features: 100 % Offline. Works without Wi-Fi or mobile data once installed. Private & Secure: Conversations stay on your device; nothing is stored in the cloud. Simple & Clean Design: Minimal interface focused on your chat experience. Why Offline AI Chat? Your messages are processed locally, making it ideal for private journaling, travel, or when you want an AI companion without worrying about data leaks or hidden costs. Perfect for Secure personal journaling Practicing conversations Creative brainstorming Fun offline companionship Take back your privacy. With Offline AI Chat, your words belong to you. Technical Details: -Built on llama.cpp: a lightweight, open‑source C++ inference engine running entirely on‑device. -Supports quantised Llama‑family models** (4‑bit, 5‑bit, 8‑bit) so the entire model fits comfortably in memory on laptops and modern smartphones. -Fully client‑side: No Python runtime, no internet calls, and no telemetry. -Automatic multi‑threading maximise performance on available hardware.