ASUS Launches Ascent QN10 AI Mini PC
ASUS has launched the Ascent QN10, a compact desktop built around Qualcomm’s Snapdragon X2 Elite platform. The company announced the system on September 17, 2026.
The QN10 combines an 18-core Qualcomm Oryon CPU with an integrated Adreno GPU. It also includes a Qualcomm Hexagon NPU capable of up to 80 TOPS of AI processing.
As a result, the system can handle on-device AI workloads without relying entirely on cloud computing. ASUS also supports agentic AI applications that can automate tasks and coordinate workflows.
Compact Hardware for Local AI
The Ascent QN10 measures less than 0.7 litres in volume. According to ASUS, it is 86% smaller than conventional 5-litre desktop systems.
Despite its size, the mini PC supports up to 32GB of LPDDR5X memory. It also offers dual M.2 SSD slots for expanded storage. Users can also connect up to four 4K displays through its HDMI 2.1 and USB4 connectivity.
The system also includes three USB4 ports and additional USB-A connections. Meanwhile, Wi-Fi 7, Bluetooth 6.0, and 2.5GbE provide wired and wireless networking options.
Power efficiency is another focus. ASUS says the Snapdragon X2 Elite configuration can deliver up to 20% faster performance while using 50% less power in specified workloads.
As a result, the QN10 is designed for sustained workloads while maintaining relatively low power consumption. ASUS also highlights quiet operation for office, studio, and enterprise environments.
AI Agents Run Directly on Device
The 80 TOPS Hexagon NPU powers the QN10’s AI capabilities. It can accelerate local AI applications, including language models, content generation, translation, video enhancement, and workflow automation.
More importantly, ASUS supports agentic AI frameworks on the device. The company lists tools including OpenClaw and Hermes Agent among supported applications.
Qualcomm also says the QN10 can run tools such as Cursor, Claude Desktop, OpenAI Codex, and OpenCode. As a result, developers can use the machine for local AI-assisted programming and agent orchestration.
The system also supports hybrid AI workflows. In this setup, local processing can handle selected workloads while cloud services manage tasks that require additional computing resources.
Developers can also access Qualcomm AI Hub for ready-to-use AI models and deployment resources. ASUS says the platform helps developers download models, deploy applications, and fine-tune workloads for the QN10.
Security and Enterprise Deployment
Security forms another part of the QN10’s design. The Snapdragon platform includes Qualcomm Secure Processing Unit technology, while Microsoft Pluton adds another hardware-based security layer.
Therefore, ASUS is positioning the mini PC for organisations that need local AI processing alongside stronger control over sensitive data. This approach can also support edge AI deployments where sending every workload to the cloud is impractical.
ASUS also says the QN10 exceeds Microsoft’s Copilot+ PC requirements. Meanwhile, its local AI capabilities allow compatible workloads to run directly on the device.
ASUS first demonstrated the QN10 at Microsoft Build 2026. At the event, Qualcomm showed the system running Visual Studio Code, GitHub Copilot, LLMWare, and AnythingLLM for local AI development.
The company has now moved the QN10 from demonstration hardware into commercial availability. In the United States, ASUS lists configurations starting at $1,349, while a 32GB version is priced at $1,699 according to current retail reporting.
Overall, the Ascent QN10 reflects the shift toward smaller computers that can run increasingly capable AI workloads locally. With Snapdragon X2 Elite, an 80 TOPS NPU, and support for agentic tools, ASUS is targeting users who want AI processing closer to their own devices.
