Nvidia Augments ChatRTX Generative AI Chatbot With Google and OpenAI LLMs, Voice Control

Nvidia has upgraded its ChatRTX generative AI chatbot with additional large language models (LLMs), including Google’s Gemma, ChatGLM3, and OpenAI’s CLIP (Contrastive Language–Image Pre-training) model. ChatRTX, available on computers with RTX GPUs, can now better search and analyze data stored locally on a PC, and users can make requests by voice as well as text.

ChatRTX LLMs

ChatRTX launched in February as “Chat with RTX” operates as a personal chatbot server run directly from a PC with an Nvidia RTX 30- or 40-series GPU. Users can upload documents and even YouTube videos, then ask the AI to search through them, explain their content, and connect them to other data using any LLM on the deviceThe interface of ChatRTX facilitates easy switching between different models, helping users choose the best option for analyzing and searching their specific data types.

Google’s Gemma model is optimized for running on high-performance laptops and desktops, like those with the appropriate Nvidia chip. is particularly notable. It aligns well with ChatRTX’s aim to simplify the use of sophisticated AI models on personal computers. Meanwhile, ChatGLM3 provides a bilingual LLM option able to communicate in both English and Chinese. OpenAI’s CLIP model rounds out the new additions to ChatRTX, offering users an AI that can sort through and parse the details of a photo collection, training the model to recognize and interpret images.

“Chatbots powered by large-language AI models have transformed computing, and NVIDIA ChatRTX lets users interact with their local data, accelerated by NVIDIA RTX-powered Windows PCs and workstations,” Nvidia product manager Jesse Clayton explained in a blog post. “ChatRTX uses retrieval-augmented generation, NVIDIA TensorRT-LLM software and NVIDIA RTX acceleration to bring chatbot capabilities to RTX-powered Windows PCs and workstations. Backed by its powerful large language models (LLMs), users can query their notes and documents with ChatRTX, which can quickly generate relevant responses, while running locally on the user’s device.”

The other major addition to ChatRTX is the introduction of voice interactivity. Nvidia has integrated the Whisper AI speech recognition system created by OpenAI into ChatRTX so that users can conduct searches and access information through voice commands. The idea is to make using the AI chatbot more convenient and accessible by making it hands-free. The ChatRTX app is available as a 36GB download from Nvidia’s website. The additional models and features for ChatRTX mark a significant step in making such powerful generative AI tools more useful, and may result in Nvidia offering its own take on ChatGPT, Claude, and other generative AI chatbots.

Follow @voicebotaiFollow @erichschwartz

Nvidia Unveils Powerful Generative AI Blackwell Chip and Platform

Poe Rolls Out ‘Price Per Message’ Generative AI Chatbot Monetization

Hugging Face Launches Free Customizable Generative AI Chatbots