MiniPC.guru
GMKtec EVO-X2

GMKtecMini PC reference

GMKtec EVO-X2

GMKtec EVO-X2 is a compact option for development, content work and local AI experiments using AMD Ryzen AI Max+ 395 and integrated Radeon 8060S graphics. Software support and the memory required by a particular model matter more than theoretical AI ratings. Memory is fixed at purchase, so choose RAM capacity for the full intended lifespan.

Processor
AMD Ryzen AI Max+ 395
Graphics
Radeon 8060S
Memory
128 GB
Storage
2 TB
Configuration being compared128 GB RAM · 2 TB SSDNew
US stores · USD

Prices across stores

MiniPC.guru compares retailer listings. Purchases are completed on the retailer’s website.

View price history

We may earn a commission if you buy through affiliate links.

RetailerAvailabilityListed priceLast checkedRetailer website
In stock$3,649.99Lowest checkedView at GMKtec
Sold by GMKtec-OfficialAffiliate link
In stock$3,649.99Lowest checkedView at Newegg
In stock$3,649.99Sold by GMKtec-USLowest checkedView at Amazon

Availability reflects the latest store check. Shipping, tax, coupons and checkout totals are not included; confirm them at the store.

Specifications

Processor, graphics and installed memory/storage describe the selected configuration. Expansion limits and connectivity describe the model; optional hardware is identified separately. References are collected in About this data below.

Processor
AMD Ryzen AI Max+ 395
Graphics
Radeon 8060S
Memory
128 GB
Storage
2 TB
Ports & displays
3 × USB-A 3.2 Gen 2, 2 × USB-A 2.0; 2 × USB4 Type-C; HDMI 2.1, DisplayPort 1.4 HBR3; SD 4.0/SDXC reader; 1 × 2.5GbE; 2 × 3.5mm audio jacks and DC input. Display output up to 7680 × 4320 at 60Hz is advertised; maximum modes depend on the selected connector and display.
Ethernet, Wi-Fi & Bluetooth
1 × 2.5GbE (Realtek RTL8125BG); Wi-Fi 7 (RZ717/MT7925); Bluetooth 5.4.
RAM & storage upgrades
Onboard LPDDR5X (non-upgradeable); Options: 64GB or 128GB; Speed: 8000MHz. Dual M.2 2280 slots; Maximum total storage: 16TB (each slot supports up to 8TB via PCIe GEN4 x4)
Power supply & power limits
External adapter: 19.5V DC, 11.8A (approximately 230W); AC input 100–240V, 50/60Hz. Supply rating, not measured wall-power consumption.
Dimensions & weight
193 × 185.8 × 77mm. Dimensions use manufacturer order; axes are not labeled.

Video

Plays on YouTube. Nothing is loaded from YouTube until you press play. YouTube

What owners say

This is a local-AI box first. With 128 GB a 70B model at Q6 fits entirely in GPU memory, but at under 4 tok/s it crawls; where the machine flies is on mixture-of-experts models, with gpt-oss-120B between 34 and 46 tok/s and Qwen3-Coder-30B-A3B between 50 and 71, while dense 32B models sit near 10. How much memory the GPU gets depends on the operating system: Windows caps it at 48 GB on the 64 GB model and 96 GB on the 128 GB one, while Linux reaches about 120 GB. It idles quiet at 27 dB(A) and becomes clearly audible under load at 47 dB(A); the fan curve is coarse and there is a P-mode button. Reviewers measured cool processors, but at least one owner saw 95-100 °C while gaming until he repasted, and the stock NVMe heatsink is widely disliked. Support is the sore point: opaque BIOS updates, machine-translated replies and months without a published security patch. It must stand vertical, has no VESA mount, runs off a 230 W external brick and offers only 2.5GbE at this price.

For whom
People who run local language models, especially mixture-of-experts ones, and want them in a single quiet-ish box.
Not for
Gaming on a budget, where a tower gives more frames per dollar, or anyone who needs ROCm on Linux to work without a fight.

Our own summary of 8 published sources, reviewed Sep 17, 2026. We have not tested this Mini PC ourselves; the sources are listed under About this data.

Performance

7 sourced resultstok/s

Same model, quantization, context length, prompt, output length, runtime version and batch size. Record time to first token separately.

Externally reported measurements. MiniPC.guru has not independently reproduced these tests.

46.05 tok/sIndependent review

GMKtec EVO-X2 Ryzen AI Max+ 395 benchmarks

By Nish Tahir · Another Dev's Two Cents

Source published
Oct 23, 2025
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 with Radeon 8060S · Radeon 8060S (gfx1151); 1 GB UMA frame buffer plus 128 GB GTT (amd_iommu=off amdgpu.gttsize=131072) · 128 GiB unified · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, ROCm/HIP backend · build 03792ad9 (6816) · Ubuntu 24.04.3 LTS, kernel 6.8.0-86-generic
Power profile
not reported by source
Test conditions
gpt-oss 120B MXFP4 MoE (59.02 GiB); tg32 at context depth 0; -fa 1 -p 2048 -n 32 -ub 2048 --mmap 0 -ngl 99; prompt processing 818.11 tok/s; ±0.18

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

34.15 tok/sIndependent review

GMKtec EVO-X2 Ryzen AI Max+ 395 benchmarks

By Nish Tahir · Another Dev's Two Cents

Source published
Oct 23, 2025
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 with Radeon 8060S · Radeon 8060S; 1 GB UMA plus 128 GB GTT · 128 GiB unified · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, ROCm/HIP backend · build 03792ad9 (6816) · Ubuntu 24.04.3 LTS, kernel 6.8.0-86-generic
Power profile
not reported by source
Test conditions
gpt-oss 120B MXFP4 MoE; tg32 with a full 32768-token context; prompt processing at the same depth 274.69 tok/s

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

65.73 tok/sIndependent review

GMKtec EVO-X2 Ryzen AI Max+ 395 benchmarks

By Nish Tahir · Another Dev's Two Cents

Source published
Oct 23, 2025
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 with Radeon 8060S · Radeon 8060S; 1 GB UMA plus 128 GB GTT · 128 GiB unified · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, ROCm/HIP backend · build 03792ad9 (6816) · Ubuntu 24.04.3 LTS
Power profile
not reported by source
Test conditions
gpt-oss 20B MXFP4 MoE (11.27 GiB); tg32 at context depth 0; -fa 1 -ub 2048 --mmap 0; prompt processing 1621.75 tok/s

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

50.16 tok/sIndependent review

GMKtec EVO-X2 Ryzen AI Max+ 395 benchmarks

By Nish Tahir · Another Dev's Two Cents

Source published
Oct 23, 2025
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 with Radeon 8060S · Radeon 8060S; 1 GB UMA plus 128 GB GTT · 128 GiB unified · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, ROCm/HIP backend · build 03792ad9 (6816) · Ubuntu 24.04.3 LTS
Power profile
not reported by source
Test conditions
Qwen3 Coder 30B A3B Q8_0 (30.25 GiB); tg32 at context depth 0; prompt processing 773.45 tok/s; the same run at 32768 tokens of context gives 27.59 tok/s

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

71 tok/sCommunity benchmark

Qwen3-Coder-30B performance benchmark on AMD Strix Halo (GMKtec EVO-X2)

By pablo-ross · GitHub pablo-ross/strix-halo-gmktec-evo-x2

Source published
Nov 4, 2025
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 with Radeon 8060S · Radeon 8060S; 1 GB video memory plus 128 GB GTT, 18.4 GB in use · 124 GiB visible of 128 GB · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-server serving the OpenAI completions API, ROCm with rocWMMA · llama.cpp b6942 (1f5accb8d), ROCm 7.0 RC · Ubuntu 24.04 LTS, kernel 6.16.9
Power profile
not reported by source
Test conditions
Qwen3-Coder-30B-A3B-Instruct Q4_K_M (17.3 GiB); -c 8192 --parallel 2 -ngl 99 --no-mmap; max_tokens 128, temperature 0.7; 51-token prompts; median of 3 runs, 71.00 ±0.02 tok/s; prompt processing 28.29 tok/s

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

22.14 tok/sCommunity benchmark

local-ai-amd-benchmark: usability benchmark for local LLMs on AMD hardware

By Kai Felix Bennett · GitHub KaiFelixBennett/local-ai-amd-benchmark

Source published
Sep 1, 2026
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 · Radeon 8060S (gfx1151); 64.4 GiB reserved to the iGPU in BIOS, 63.6 GiB visible to Windows · 128 GiB unified LPDDR5X · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, Vulkan backend, no speculative decoding · build 580e88d · Windows 11 Pro 26200, Adrenalin 32.0.31021.5001, Vulkan SDK 1.4.350.0
Power profile
not reported by source
Test conditions
Qwen3.8-Flash-Next UD-Q4_K_XL (177B total, about 6B active); token generation at 512 tokens of context depth; median of 3 runs

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

8.84 tok/sCommunity benchmark

local-ai-amd-benchmark: usability benchmark for local LLMs on AMD hardware

By Kai Felix Bennett · GitHub KaiFelixBennett/local-ai-amd-benchmark

Source published
Sep 1, 2026
Test performed
Not reported by source
Added to MiniPC.guru
Sep 17, 2026
Source reviewed
Sep 17, 2026
Tested hardware
AMD Ryzen AI Max+ 395 · Radeon 8060S; 64.4 GiB reserved to the iGPU in BIOS · 128 GiB unified LPDDR5X · not applicable: the source does not identify the installed SSD, and token generation does not read from it once the model is resident in memory
Software / version / OS
llama.cpp llama-bench, Vulkan backend, no speculative decoding · build 580e88d · Windows 11 Pro 26200
Power profile
not reported by source
Test conditions
Qwen3.8-Flash-Next UD-Q4_K_XL; token generation at 131072 tokens of context depth; single run, not a median

Editorial review checks the source and documentation; it does not certify the measurement or mean we tested the device ourselves.

Price history

GMKtec EVO-X2 · 128 GB RAM · 2 TB SSD

Loading checks…

New condition · US stores · USD · Selected configuration only. Prices exclude shipping and tax. Each price stays on the chart until that listing's next check, for at most 36 hours, the same window the store prices use. Out-of-stock, failed and missing checks leave gaps.

About this data10 references

Specs and any benchmark results come from the sources below. We haven't tested this Mini PC ourselves.

Sources

  1. GMKtec EVO-X2 — manufacturer specifications ↗
    • Manufacturer / specification source
    • Ports & displays · Reviewed Sep 12, 2026
    • Ethernet, Wi-Fi & Bluetooth · Reviewed Sep 12, 2026
    • RAM & storage upgrades · Reviewed Sep 12, 2026
    • Power supply & power limits · Reviewed Sep 12, 2026
    • Dimensions & weight · Reviewed Sep 12, 2026
  2. Selected configuration ↗
    • Selected configuration
  3. GMKtec EVO-X2 Ryzen AI Max+ 395 benchmarks ↗
    • External benchmark · Reviewed Sep 17, 2026
    • Owner review source · Reviewed Sep 17, 2026
  4. Qwen3-Coder-30B performance benchmark on AMD Strix Halo (GMKtec EVO-X2) ↗
    • External benchmark · Reviewed Sep 17, 2026
    • Owner review source · Reviewed Sep 17, 2026
  5. local-ai-amd-benchmark: usability benchmark for local LLMs on AMD hardware ↗
    • External benchmark · Reviewed Sep 17, 2026
    • Owner review source · Reviewed Sep 17, 2026
  6. One of the best mini PCs of 2025: AMD Ryzen AI Max+ 395 and Radeon 8060S reviewed in the GMKtec EVO-X2 ↗
    • Owner review source · Reviewed Sep 17, 2026
  7. GMKtec EVO-X2 AI mini PC review: compact Strix Halo powerhouse ↗
    • Owner review source · Reviewed Sep 17, 2026
  8. GMKtec EVO-X2 review: an AMD Ryzen AI Max+ 395 powerhouse ↗
    • Owner review source · Reviewed Sep 17, 2026
  9. GMKtec EVO-X2 review: a mini PC that packs serious AI and gaming power ↗
    • Owner review source · Reviewed Sep 17, 2026
  10. GMKtec EVO-X2 Ryzen AI Max+ 395 owners thread ↗
    • Owner review source · Reviewed Sep 17, 2026

Related models

Models with related processors, graphics or use cases. Check each model’s specifications and configurations before comparing.

Beelink GTR9 ProAMD Ryzen AI Max+ 395GEEKOM A9 Mega (Max+ 395)AMD Ryzen AI Max+ 395GMKtec EVO-X3AMD Ryzen AI Max+ 395MINISFORUM MS-S1 MAXAMD Ryzen AI Max+ 395