I kept bouncing between ChatGPT, Gemini, and Grok for stuff that didn't need a massive model and all of them wanted my data or my money. n0x runs in your browser LLM inference on your GPU, document Q&A, web search, Python execution, image gen, memory. Pick a model, start using it. Nothing leaves your machine. Works offline after first visit. Supports Ollama and any OpenAI-compatible API for bigger models.