Ask questions across your notes and the web, then open the note, heading, PDF page, or URL behind every answer. Use local models with Ollama or LM Studio, or connect a cloud API.