ለዘላለም በነፃ ከኢንተርኔት ውጭ የሚሰራ AI (በስልክ እና በላፕቶፕ)

In this video, we cover how to run 100% Free, Offline AI models locally on your smartphone and PC without an internet connection! https://abyssiniya.ai/ We walk step-by-step through setting up Ollama and running top open-source AI models like DeepSeek-R1, Google Gemma 4, Meta Llama 3.2, and Qwen 2.5. We’ll also break down the exact hardware specifications (RAM, VRAM, and Processors) required for both Android phones and laptops/PCs. 📌 What You'll Learn in This Video: Why Local AI? Complete privacy, zero subscription fees, and no internet/data required. Top Models Explained: DeepSeek-R1 (Reasoning), Gemma 4 (Multimodal & Fast), Llama 3.2 (Lightweight), and Qwen (Coding). Ollama Installation: Quick setup guide for macOS, Windows, Linux, and Android. Hardware Benchmarks: Exactly what RAM & GPU you need for 1.5B, 3B, 8B, and 14B models. 🤖 Popular Models Featured in This Guide: DeepSeek-R1 (deepseek-r1:8b / 1.5b): Industry-leading open-weights reasoning model for math, coding, and logic. Google Gemma 4 (gemma4): Google's newest lightweight, highly efficient multimodal open model. Meta Llama 3.2 (llama3.2:1b / 3b): Ultra-fast, lightweight models tailored for mobile devices. Alibaba Qwen 2.5 (qwen2.5 / qwen2.5-coder): Powerful coding and general reasoning capabilities. 💻 System Requirements / Recommended Specs: 📱 Mobile / Android Specs: Minimum (1B – 3B Models): 6GB RAM + 8GB Free Storage (Runs Llama 3.2 1B / Qwen 1.5B) Recommended (3B – 8B Models): 8GB–12GB+ RAM | Snapdragon 8 Series / Dimensity 8000+ 🖥️ PC & Laptop Specs: Minimum (CPU Only): 8GB RAM | 4-Core CPU | 15GB Free Storage (For 1.5B – 3B models) Recommended (Smooth Performance): 16GB–32GB RAM | 8GB VRAM GPU (Nvidia RTX 3060/4060 or Apple Silicon M1/M2/M3) Power User (Heavy Reasoning Models): 32GB–64GB RAM | 12GB+ VRAM (Runs 14B – 32B models effortlessly)