Computing

NVIDIA RTX Spark Laptops Gain New AI Tools and Models

NVIDIA has announced a slate of updates for its thin and light RTX Spark Windows laptops, introducing new models, agent tools, and enhanced gaming capabilities.

Radwan AbdulhadiEditor and technology writer

Published: October 28, 2026 · Updated: October 28, 2026 · 3 min read

NVIDIA RTX Spark Laptops Gain New AI Tools and Models — Image: NVIDIA Blog

New RTX Spark Laptops and Superchip Architecture

NVIDIA RTX Spark powers thin and light laptops built for developers, creators, and gamers alike. These systems are now available from major manufacturers including ASUS, Dell, HP, Lenovo, Microsoft, and MSI. Built around an advanced hardware architecture, these powerful laptops combine an NVIDIA Blackwell RTX GPU and an NVIDIA Grace CPU into a single superchip equipped with up to 128GB of unified memory and delivering up to 1 petaflop of FP4 AI performance.

By integrating the NVIDIA CUDA platform and RTX graphics onto a single power-efficient superchip, these premium Windows laptops achieve designs as thin as 14mm and as light as three pounds while offering all-day battery life. Because creative projects and local AI models draw from the same unified memory pool, these portable machines can handle demanding workloads previously reserved for desktop workstations, even when operating on battery power.

Advanced Speech Transcription with Nemotron 3.5

Among the newly available software updates for the platform is NVIDIA Nemotron 3.5 Transcribe, a promptable open-weight speech model designed to convert recorded conversations into structured information. Rather than requiring separate models for basic transcription, speaker attribution, and translation, this single model handles all three simultaneously while supporting 40 languages, including seven Indic languages.

Developers can guide the speech model using plain-language instructions to summarize meeting decisions or incorporate specialized vocabulary for fields like healthcare and financial services. According to evaluation data, Nemotron 3.5 Transcribe delivers high-tier accuracy with approximately 4.3 billion active parameters, making intelligent transcription practical for deployment on local hardware.

Resource Management and Liquid AI Integration

As users run long-term AI agents alongside traditional applications on their RTX Spark devices, balancing system compute becomes crucial. A new resource management feature integrated into the NVIDIA App allows users to choose how GPU resources are shared, letting them prioritize active foreground applications for maximum performance or allocate additional capacity to background-running agents.

Additionally, the platform gains day-zero support for Liquid AI's LFM3 family of Liquid Foundation Models. Spanning text, vision, audio, and decision models built on a shared backbone, these compact models support local workflows such as automated inbox triage, calendar management, and document reading directly on RTX Spark systems.

Enhanced Gaming and RTX Remix Updates

On the gaming front, S-GAME's action role-playing title Phantom Blade Zero launches with advanced ray tracing and NVIDIA DLSS support. Players can experience enhanced bounced lighting, detailed reflections, and realistic shadows on RTX Spark laptops and other GeForce RTX-powered hardware.

Furthermore, the NVIDIA RTX Remix modding platform update brings path tracing and DLSS capabilities to Minecraft Java Edition 1.21.1. Portal with RTX and the Half-Life 2 RTX demo also receive native RTX Spark support, introducing 3D-Guided Neural Rendering and optimized performance modes for frame generation.