Nvidia updates ChatRTX with new large language models and more

Back in February, we got our first look at Nvidia's ChatRTX, an offline AI chatbot that can help users sift through files and data on their PCs using Retrieval Augmented Generation, in conjunction with the TensorRT Large Language Model, which users can customise themselves by granting the app access to specific folders and directories on your system. Now, ChatRTX has been updated, bringing support for new AI models, as well as voice and image functionality. 

Chat with RTX, now known as simply ‘ChatRTX' had access to two AI models at launch, Mistral and Llama. Now after Nvidia's latest update, the software now supports more models, including Google's Gemma LLM, as well as the latest versions of Llama and Mistral.

Moreover, ChatRTX now allows users to search for photos in local directories using CLIP (Contrastive Language-Image Pre-training), an OpenAI model. On the voice side, ChatRTX can now recognise voice input in multiple languages.

Users should note that the software requires 35GB of storage (the installer alone is 11GB). ChatRTX currently only works on GeForce RTX 30 and RTX 40 GPUs with at least 8GB of VRAM and 16GB of system memory.

Discuss on our Facebook page, HERE.

KitGuru says: Do you prefer local AI assistants over online ones? Have you already tried ChatRTX?

Become a Patron!