¡Adiós Claude! IA GRATIS y en LOCAL: Uso QWEN 35B A3B

In this video, we use one of Qwen's latest AI LLM models: Qwen 3.6 35B A3B, which performs very well even on home hardware. In my case, I'm using an RTX 3060 GPU (12GB) + 64GB of RAM + AMD Ryzer 7 5800X, and I'm achieving speeds of 50 tokens/second. This model can also be used in programs like Opencode, to function as an agent. TIMELINE: 00:00 Ollama vs LM Studio vs Llama.cpp 01:55 Local Artificial Intelligence 03:48 The Llama.cpp Wrapper 08:59 Testing Gemma4 E2B 12:19 Testing Qwen3.6 35B A3B If anyone is interested in the script, I've renamed it "shama" and uploaded it to GitHub, along with detailed instructions for installing llama.cpp and more: https://github.com/manzdev/shama LINKS: 📕 Ollama: Local Models https://lenguajejs.com/aprender-a-pro... 🎬 Artificial Intelligence Playlist:    • OpenCode + Ollama (Inteligencia Artificial)   My social media:   / manz     / manzdev     / joseromanhdez     / manz.dev   ⚡ Other social media: https://links.manz.dev/