Enter the base URL for your local LLM server (e.g., LM Studio, Ollama)
Select the vision-capable model to use for image analysis