Sanctum AI

Sanctum AI supports manual model updates. You'll need to download the desired open-source model files (e.g., weights and configuration) from sources like Hugging Face and import them into Sanctum AI. The application likely provides a mechanism for specifying the model's location and configuration, but you will need to handle the downloading and initial setup yourself.

While Sanctum AI offers a chat interface, its core functionality is running local LLMs. You can use it for various tasks beyond chat. You could process text files by feeding them into the LLM through the chat interface, or integrate it into your local development workflow by writing scripts that interact with the loaded model via an API (if available, and if the selected model supports such tasks).

Sanctum AI is built with a focus on complete data sovereignty and privacy. Because it runs locally and offline, there are no built-in logging or telemetry features that send data externally. However, it's always good practice to review the application's documentation or settings to confirm that no optional features are enabled that might inadvertently collect data.

As a relatively new, open-source tool, community resources for Sanctum AI may be limited. Check the Sanctum AI website on Futurepedia for links to official documentation or community forums, or search for relevant discussions on platforms like GitHub or Reddit. Since it supports popular models like Llama and Mistral, resources for those models may also be helpful.