Nvidia Adds 40-Language Speech Model and GPU Controls to RTX Spark

Nvidia expanded software support for its RTX Spark laptops Thursday, adding a 40-language speech model and tools to distribute GPU capacity between applications and background AI agents.

The update supports Nemotron 3.5 Transcribe, an open-weight model for transcribing speech, identifying speakers and translating audio. Liquid AI’s LFM3 models are also supported and can run locally and be fine-tuned.

Separate GPU controls let users allocate computing capacity between applications in the foreground and AI agents in the background.

RTX Spark laptops have been launched by vendors including ASUS, Dell, HP, Lenovo, Microsoft and MSI. They start at $2,599, with configurations offering up to 128 gigabytes of unified memory and 1 petaflop of FP4 AI performance.