Nvidia Launches "Chat with RTX" Free on Device AI Models

AI Capabilities

Nvidia's latest innovation, "Chat with RTX," leverages the open-source AI model Mistral to manage various file formats directly on personal computers. This tool supports formats like PDF, Doc, DocX, and XML, and even incorporates YouTube playlist transcriptions to enhance search capabilities 1. Jaeden Schafer highlights the model's ability to extract information from files by simply typing questions, making it easier to locate specific content 1.

This lets users extract information from their files by typing in questions, right? So it's not even just like, where is a specific file? It's like, hey, I talked about this in a file. Where is that?

--- Jaeden Schafer

However, the tool requires significant storage, between 50 and 100GB, to accommodate its operations 1.

   

Limitations

Despite its advanced capabilities, "Chat with RTX" has limitations, such as the inability to remember context between questions, which affects response relevance 2. Jaeden Schafer explains that unlike chat GPT, which can maintain a conversation, this tool operates on a single-shot question basis 2.

Rather than searching through notes or saved content, users can simply type queries.

--- Jaeden Schafer

Running AI models locally, like "Chat with RTX," offers privacy benefits and cost-effectiveness compared to cloud-hosted alternatives, aligning with a trend towards offline AI model usage 2.