Skip to main content

RTX Spark Surface Laptop leads reported Qwen agent benchmark

17 OCTOBER 2026·2 MIN READ·6 SOURCES

Tom’s Hardware reported that Microsoft’s RTX Spark-powered Surface Laptop Ultra outperformed Apple’s M5 Max MacBook Pro by about 25% in agentic AI completion benchmarks using Qwen models.

RTX Spark Surface Laptop leads reported Qwen agent benchmark

Key takeaways · 4

  • 01

    Tom’s Hardware reported an approximately 25% lead for the Surface Laptop Ultra in the tests.

  • 02

    The comparison measured agentic AI completion and used Qwen models.

  • 03

    Nvidia says RTX Spark combines a Blackwell RTX GPU and Grace CPU, with up to 128GB of unified memory.

  • 04

    Nvidia said RTX Spark laptops were available from ASUS, Dell, HP, Lenovo, Microsoft and MSI.

What the benchmark shows

Tom’s Hardware reported that Microsoft’s Surface Laptop Ultra, powered by Nvidia’s RTX Spark, beat Apple’s M5 Max MacBook Pro by about 25% in the reported tests.[1] The comparison focused on agentic AI completion benchmarks and used Qwen models.[1] That makes the result relevant to teams evaluating local AI-agent work on laptops, but it should be read within the scope of those tests. The evidence does not establish that the Surface is faster across other models, applications or kinds of computing.

Nvidia’s hardware claims

Nvidia unveiled RTX Spark as a superchip designed for Windows PCs and personal AI agents.[2] The company says it combines a Blackwell RTX GPU with a Grace CPU, provides up to one petaflop of FP4 AI performance and supports up to 128GB of unified memory.[3] Nvidia also says the hardware can run large language models and AI agents locally.[3] These are vendor specifications and capability claims; the benchmark report supplied here does not show how each contributed to the reported result.

Local agents and controls

Nvidia says RTX Spark PCs can run large language models and AI agents locally, and lists GitHub Copilot, Hermes, OpenClaw, Perplexity, Pi and WorkBuddy in its agent setup.[3] Nvidia also says its OpenShell runtime and Microsoft Execution Containers help users control agent permissions.[3] The company describes RTX Spark as capable of running 120-billion-parameter models with up to a one-million-token context window using agents locally.[2] For teams assessing local workflows, agent compatibility and permission controls are relevant alongside benchmark speed.

Availability and limits

Nvidia said RTX Spark laptops were available from ASUS, Dell, HP, Lenovo, Microsoft and MSI, while Tom’s Hardware reported a starting price of $2,599 for the Surface Laptop Ultra with RTX Spark.[3][6] The price is a useful point of comparison for buyers, but the supplied evidence gives no broader pricing or configuration comparison between the tested laptops. Tom’s Hardware characterized the Windows-on-Arm platform as strong in local AI but weak for gaming.[4] Nvidia, by contrast, says RTX Spark PCs can play AAA games at 1440p and more than 100 frames per second.[3]

For professionals choosing laptop hardware for local AI agents, the report offers a concrete benchmark result, but only for the stated Qwen-based tests. Buyers should weigh it alongside price, workload fit, agent controls and the conflicting descriptions of gaming capability.

Why it matters
Story quiz

Test yourself on this story — 1 question.

Create a free account to take the quiz, earn XP, and get a daily session built for your industry.

Take the quiz

How this developed

  1. 17 October 2026

    RTX Spark Surface Laptop leads reported Qwen agent benchmark

Sources

AI fluency, one session a day, built for your work.