Use Local LLM extension: run llm locally (LLama 70B or DeepSeek with WebLLM + Gemini Nano), ask ai models on your tabs - private ai.
