Beelink SER10 MAX AMD Pro Ryzen™ AI 9 HX 470 (OpenClaw & Local LLM Pre-installed)
Price range: $1,339.00 through $2,299.00
12 Zen5 CPU cores, combined with a new Radeon™ 890M iGPU, a new generation XDNA 2 NPU and 86 AI TOPS, the Ryzen™ AI 9 HX470 processor’s fully upgraded architecture significantly boosts the SER10 MAX’s overall performance, delivering a faster, smarter, and more secure Al PC experience!.
The Beelink SER10 MAX is built around AMD’s newest Ryzen AI 9 HX 470 — a 12-core hybrid chip pairing four full-power Zen 5 cores with eight efficiency-focused Zen 5c cores, backed by an XDNA 2 NPU for up to 86 TOPS of combined AI performance. It’s a compact machine aimed at buyers who want strong everyday performance, fast wired networking, and genuine on-device AI capability, with OpenClaw and local LLM tooling pre-installed and ready to run from first boot.
Key Features
- AMD Ryzen™ AI 9 HX 470 — 12 Cores, 24 Threads — A hybrid design combining 4 Zen 5 performance cores with 8 Zen 5c efficiency cores on a 4nm process, boosting up to 5.2GHz with 36MB of combined L2+L3 cache — strong single-thread speed for responsive everyday use and enough parallel throughput for heavier multitasking and compilation.
- AMD Radeon™ 890M Graphics, RDNA 3.5 — A 16-compute-unit iGPU running at up to 3100MHz delivers smooth 4K output and handles light-to-moderate gaming and GPU-accelerated tasks without a discrete card.
- Up to 86 AI TOPS via XDNA 2 NPU — Combined CPU, GPU, and dedicated NPU throughput reaches up to 86 TOPS, giving OpenClaw and other AI workloads real on-device acceleration for local inference, not just cloud-dependent processing.
- 32GB or 64GB DDR5 5600MT/s Memory — Dual-channel DDR5 in 32GB or 64GB configurations gives you room for large browser sessions, VMs, and containerized dev environments running alongside AI workloads.
- 10Gbps Wired Networking — A 10GbE port delivers four times the throughput of standard 2.5GbE, ideal for fast backups, large file transfers, and low-latency connections that eliminate the bottlenecks typical mini PC networking runs into.
- Up to 8TB of PCIe 4.0 NVMe Storage — Dual M.2 2280 PCIe 4.0 slots support up to 8TB combined (4TB per slot), enough room for a large OS/apps drive plus a dedicated volume for models and projects.
- MSC 2.0 Vapor Chamber Cooling at 32dB — Beelink’s vapor chamber, quiet-fan, and metal dust filter combination keeps the HX 470 stable under sustained load while staying near-silent — a real consideration if this machine lives on a desk next to you all day.
- USB4 (40Gbps) and 4K Display Output — A full-function USB4 port plus HDMI 2.1 and DisplayPort support high-resolution multi-monitor setups for both productivity and media.
Why Choose the SER10 MAX
The SER10 MAX sits at a genuinely useful spot in AMD’s current Ryzen AI lineup: a 12-core hybrid CPU with real efficiency cores for background tasks, a capable iGPU, and an NPU that pushes the platform’s combined AI throughput to 86 TOPS — enough headroom for OpenClaw to run meaningful local inference rather than relying entirely on cloud API calls. Combined with 10Gbps Ethernet, it’s also a strong pick for anyone building a home lab or small office setup that needs fast wired networking alongside AI capability.
For developers experimenting with agentic coding tools, the appeal is being able to run smaller quantized models locally through the NPU-accelerated pipeline for latency-sensitive or privacy-sensitive tasks, while still having the CPU and GPU muscle to handle everything else a daily workstation needs — compiling, testing, multitasking across dozens of tabs and tools.
If your workloads call for the largest possible local models, our GTR9 Pro with its 128GB unified memory pool will get you further; the SER10 MAX’s strength is being a well-rounded, quiet, fast everyday machine with genuine AI acceleration built in, at a more accessible footprint and price. Every SER10 MAX ships from our UK warehouse with 1-3 business day processing, free shipping over $199, and ComputaHardware support if you need help setting up OpenClaw or your local model stack.
Specifications
- CPU: AMD Ryzen™ AI 9 HX 470, 12C/24T (4x Zen 5 + 8x Zen 5c), up to 5.2GHz boost, 36MB cache, 4nm
- GPU: AMD Radeon™ 890M, 16 CUs, up to 3100MHz, RDNA 3.5
- AI: XDNA 2 NPU, up to 86 combined AI TOPS
- Memory: 32GB or 64GB dual-channel DDR5 5600MT/s
- Storage: Dual M.2 2280 PCIe 4.0 slots, up to 8TB combined
- Ports: USB4 (40Gbps/DP1.4), HDMI 2.1, DisplayPort
- Networking: 10Gbps Ethernet, Wi-Fi 6 (Intel AX200), Bluetooth 5.2
- Cooling: MSC 2.0 vapor chamber, ~32dB near-silent operation
- Colors: Frost Silver, Space Gray
- Software: OpenClaw and local LLM tooling pre-installed; Windows 11 Pro or Ubuntu
| RAM+SSD | 32GB(16G*2) DDR5 5600 MT/s + 1TB SSD, 64GB (32G*2)DDR5 5600 MT/s + 2TB SSD (Pre-order), 64GB(32G*2) DDR5 5600 MT/s + 1TB SSD (Pre-order), 96GB(48G*2) DDR5 5600 MT/s + 2TB SSD (Pre-order) |
|---|---|
| Operating System | Ubuntu, Ubuntu + Windows |
| Power Adapter | AU, EU, JP, UK, US |
6 reviews for Beelink SER10 MAX AMD Pro Ryzen™ AI 9 HX 470 (OpenClaw & Local LLM Pre-installed)
Fast, Tracked Worldwide Shipping
In-stock orders are processed within 1–3 business days and shipped with full tracking. Delivery costs and timeframes depend on your destination:
Delivery Estimates & Costs
- United Kingdom — 2–4 business days, $9.95 (free over $199)
- Europe — 5–10 business days, $19.95 (free over $199)
- United States & Canada — 6–12 business days, $24.95 (free over $199)
- Rest of world — 10–20 business days, $49.95 (free over $499)
Free shipping is applied automatically at checkout once your order passes the threshold for your destination — no code needed. Every order ships with a tracking number emailed to you as soon as it leaves our warehouse.
International orders may be subject to import duties or customs fees set by your local authorities; these are not included in the checkout price and are the buyer’s responsibility.
Read our full Shipping Policy for details on processing, customs, damaged or lost parcels, and delivery addresses.
Related products
Beelink EQR5 AMD Ryzen™ 5 5500U
Price range: $409.00 through $509.00
Beelink EQi13 Intel® Core™ 13500H/13420H
Price range: $489.00 through $599.00
Beelink SER8 AMD Ryzen™ 7 8745HS
Price range: $639.00 through $889.00
Beelink SER10 MAX AMD Pro Ryzen™ AI 9 HX 470
Price range: $1,299.00 through $1,599.00
Beelink SER9 MAX AMD Ryzen™ 7 H 255/H260/H350
Price range: $899.00 through $1,289.00
Beelink EQR7 AMD Ryzen™ 7 7735HS/7735U
Price range: $479.00 through $589.00
Beelink GTi15 Ultra Intel® Core™ Ultra 9 285H
Price range: $1,499.00 through $1,949.00
Nia J. –
Performance is there but I had some early instability with the machine randomly restarting under heavy sustained load, seemed to resolve itself after a BIOS update from Beelink’s site. A bit of a rough start but stable for the past few weeks now.
Adam S. –
This has become my main machine for both work and homelab tinkering. The 10GbE networking paired with the AI capability makes it a genuinely unique product in this size. Runs cool, quiet, and the OpenClaw pre-install saved me a lot of setup time.
Rosa M. –
Solid all rounder. The 64GB RAM variant handles my dev environment, several containers and a local model comfortably. Only complaint is it took about ten days to arrive, a bit longer than the estimate given at checkout, but the product itself has been great.
Katarzyna W. –
Really impressed with the build quality and thermal performance. I moved from a much older AMD mini PC and the difference in responsiveness is massive. Only wish the box included a slightly better HDMI cable, had to swap mine out for a higher spec one for 4K at higher refresh.
Finn O. –
The 10Gbps LAN port alone justifies this purchase for my home server setup. Running it with OpenClaw and a local model for some agentic coding experiments, and the AI TOPS on offer is plenty for the smaller to mid-size models I use day to day. Vapor chamber cooling keeps it dead quiet.
Xavier D. –
Fantastic little AI workstation. Local LLM setup was mostly pre-configured, just had to download the models I wanted through Ollama and it was working within the hour. Fan noise under sustained inference load is noticeably lower than I expected for the performance on offer.