NVIDIA has introduced Chat with RTX, a local chatbot that runs directly on your PC, providing users with the ability to access their files and documents for personalized and contextually aware interactions. Unlike traditional chatbots that often rely on cloud processing, Chat with RTX allows users to process sensitive data locally, enhancing both safety and contextual awareness.
The chatbot opens the door to a unique experience where users can feed it a selection of personal data, enabling it to generate summaries based on the provided information. The program supports receiving data from various types of files, such as .txt, .pdf, .doc/.docx, and .xml. The process of loading relevant files into its dataset is seamless, taking only seconds.
One standout feature of Chat with RTX is its integration with YouTube videos and playlists. Users can add video URLs to the dataset, allowing the chatbot to extract knowledge from the content for contextual queries. This opens up possibilities for seeking travel advice based on favorite influencer videos or obtaining tutorials and summaries derived from educational resources.
NVIDIA envisions its AU chatbot as a step toward a more personalized digital assistant that operates within the contextual framework of users’ personal data. Unlike most chatbots that send data to the cloud, this local chatbot ensures that sensitive information remains on the user’s PC, eliminating the need to share it with third parties or rely on an internet connection.
Despite its innovative features, Chat with RTX comes with certain limitations. As a demo product, users should expect some bugs, which NVIDIA plans to address based on user feedback and error reports. Additionally, strict hardware requirements narrow the availability of Chat with RTX, limiting its functionality to Windows PCs equipped with NVIDIA GeForce RTX 30 Series GPUs or higher and at least 8GB of VRAM.
