Skip to product information
New
MINISFORUM MS-01 Workstation MINISFORUM US
  • MINISFORUM MS-01 Workstation MINISFORUM US
  • MINISFORUM MS-03
  • MINISFORUM MS-03
  • MINISFORUM MS-03
  • MINISFORUM MS-03
  • MINISFORUM MS-03
  • MINISFORUM MS-03

Free Shipping

Free Shipping

Free RMA return

Free RMA return

0% APR plan

0% APR plan

2-Year Seller Warranty

2-Year Seller Warranty

30-Day Money-Back Guarantee

30-Day Money-Back Guarantee
MINISFORUM US

MINISFORUM MS-03

$999.00 $1,249.00

For bulk discount inquiries, please contact us at elsa@minisforum.com

  • Optional Intel® Core™ Ultra 9 386H | Intel® Core™ Ultra 5 336H
  • Intel® Graphics
  • Fluent 35B Local LLMs(Gemma 22.05 tok/s) | NPU Up to 50
  • vPro Remote Mgmt | Out-of-band IT control
  • 2.5GbE & 10GbE | WiFi 7 & BT 6
CPU:

Intel® Core™ Ultra 9 386H

Type:

Barebone | Estimated shipping time: Late August

Adapter:

US

Payment Methods:

American Express Apple Pay Cartes Bancaires Diners Club Discover Google Pay JCB Maestro Mastercard PayPal Union Pay Visa Visa Electron

Trusted Delivery:

Trusted Delivery:

Free Shipping

Free Shipping

Free RMA Return

Free RMA Return

0% APR Plan

0% APR Plan

2-Year Seller Warranty

2-Year Seller Warranty

30-Day Money-Back Guarantee

30-Day Money-Back Guarantee

Compare Specs

$999.00 $1,249.00 Save $250.00
Buy Now
Purchase
Overview
Specs
Reviews

MS-03 Edge AI Mini Workstation

Local AI. Remote Control

Run 35B models locally. Push Gemma 26B at 22.05 tok/s. Connect with dual 10G SFP+. Manage remotely with vPro. Store more with U.2 enterprise SSD support. Compact in size. Built for edge AI deployment.

MS-03 Edge AI Mini Workstation

MS-03 supports vPro remote management on select SKUs, making it ready for retail sites, branch offices, labs, and remote teams. Devices stay on-site. IT stays in control.

Run Qwen, Gemma, GPT-OSS, and more locally. Test quantized models, build low-latency AI assistants, deploy private knowledge bases, and run inference services on your own hardware.

Keep contracts, code, medical files, legal documents, and business knowledge bases off the cloud. Models, data, and inference workflows stay inside your own environment.

Dual 10G SFP+, 10G RJ45, 2.5G RJ45, and flexible storage expansion make MS-03 a strong fit for edge computing, routing, high-speed NAS, data preprocessing, and local AI nodes.

Measured Gains
Faster Workflows
Intel® Core™ Ultra 9 386H
Intel® Core™ Ultra Series 3 · Panther Lake
More Speed for Demanding Workflows In Cinebench R23, MS-03 scores 2,111 in single-core and 21,149 in multi-core performance-up 16.9% and 30.5% over the Intel® Core™ i9-13900H in MS-01. More speed for local AI, rendering, code compilation, and heavy multitasking.
Intel® Graphics
A Major Leap in Integrated Graphics The compute upgrade carries into graphics. MS-03 scores 3,290 in 3DMark Time Spy, delivering 77.45% higher graphics performance than the Intel® Iris® Xe Graphics in the Core™ i9-13900H. Hardware ray tracing, AV1 encode/decode, and XeSS support bring smoother video, content creation, and light 3D workflows.
*Performance comparisons are based on MINISFORUM laboratory testing of the Intel® Core™ Ultra 9 386H in MS-03 and the Intel® Core™ i9-13900H in MS-01. Actual results may vary by system configuration, power mode, cooling conditions, software version, and test environment.

Deploy Local AI Where Work Happens

Private Knowledge Base

Deploy business documents, code repositories, contracts, and training files locally. MS-03 hosts models, vector databases, and private datasets, keeping AI search and summarization inside your network.

Branch AI Endpoint

Deploy MS-03 in retail sites, branch offices, or remote workplaces to bring local AI to the edge. With vPro remote management on select SKUs, IT can reduce on-site service calls.

Local Coding Assistant

Run an AI coding assistant inside your LAN for code explanation, script generation, project summaries, and internal documentation Q&A. Keep code in-house. Bring AI into the workflow.

High-Speed Edge Data Processing

Dual 10G SFP+ connects directly to switches, NAS systems, or servers, enabling fast data ingestion, preprocessing, analysis, and local AI response.

Built for Local AI

Across real model tests, MS-03 delivers smooth local inference with Qwen 35B, Gemma 26B, and GPT-OSS 20B.

Up to 128GB DDR5 7200 gives large models, long-context workloads, multitasking, and private knowledge-base services more room to run.

U.2 enterprise SSD support is built for large model files, vector databases, and 24/7 high-read workloads. Local AI needs speed, capacity, and reliable access.

Dual heatpipes, a blower-style fan, and phase-change material help maintain stable performance across sustained inference, edge services, and always-on local AI workloads.

Enterprise StorageFlexible Upgrade

Enterprise StorageFlexible Upgrade

U.2 NVMe Support

U.2 enterprise SSD support via adapter gives large model files, vector databases, and high-read workloads a stronger local storage foundation.

3 × M.2 SSD (2×PCIe 5.0 + 1×PCIe 4.0)

Three M.2 SSD slots—two PCIe 5.0 and one PCIe 4.0—let you separate the OS, models, and datasets for cleaner local AI workflows and easier upgrades.

PCIe x16 Physical Slot

Supports low-profile add-in cards for capture, networking, and specialized deployments.

On-Site Devices

Remote IT Control

Select MS-03 SKUs support Intel vPro for remote maintenance, diagnostics, and recovery. For distributed AI nodes, IT does not need to be on-site for every issue.

Sustained Cooling for
Stable Inference

Dual-Heatpipe Cooling

Dual heatpipes rapidly move heat away from the
core, helping reduce performance fluctuations
under sustained workloads.

Dual heatpipes rapidly move heat away from the core, helping reduce performance fluctuations under sustained workloads.

Blower-Style Fan

Focused airflow exhausts heat efficiently, helping
the compact chassis sustain performance.

Focused airflow exhausts heat efficiently, helping the compact chassis sustain performance.

Built for Long AI Workloads

Ready for local model inference, edge services,
knowledge-base search, and 24/7 workloads.

Ready for local model inference, edge services, knowledge-base search, and 24/7 workloads.

Dual-Heatpipe Cooling

Dual heatpipes rapidly move heat away from the
core, helping reduce performance fluctuations
under sustained workloads.

Dual heatpipes rapidly move heat away from the core, helping reduce performance fluctuations under sustained workloads.

Blower-Style Fan

Focused airflow exhausts heat efficiently, helping
the compact chassis sustain performance.

Focused airflow exhausts heat efficiently, helping the compact chassis sustain performance.

Built for Long AI Workloads

Ready for local model inference, edge services,
knowledge-base search, and 24/7 workloads.

Ready for local model inference, edge services, knowledge-base search, and 24/7 workloads.

Full Connectivity and Expansion

Packing List

MS-03

AC Power Adapter

Power Cord

HDMI™ Cable

SSD Heatsink

Customized U.2 to M.2

U.2 Mounting Screw Set

User Manual

MS-03 Specification

Model

MS-03-S386H

MS-03-S336H

CPU

Processor: Intel® Core™ Ultra 9 386H(vPro Enabled), 16C/16T

Efficient-core Base Frequency: 1.6 GHz

Max Turbo Frequency: 4.9 GHz

TDP (Balance Mode): PL1 70W, PL2 80W

TDP (Quiet Mode): PL1 45W, PL2 45W

Processor: Intel® Core™ Ultra 5 336H (vPro Enabled), 12C/12T

Efficient-core Base Frequency: 1.5 GHz

Max Turbo Frequency: 4.6 GHz

TDP (Balance Mode): PL1 65W, PL2 65W

TDP (Quiet Mode): PL1 45W, PL2 45W

GPU

Graphics Cards: Intel® Graphics

Max Dynamic Frequency: 2.5 GHz

GPU Peak TOPS (Int8): 40

Graphics Cards: Intel® Graphics

Max Dynamic Frequency: 2.3 GHz

GPU Peak TOPS (Int8): 37

NPU

NPU Peak TOPS (Int8): 50

NPU Peak TOPS (Int8): 47

Memory

Memory Type: DDR5 (SO-DIMM Slot х 2)
Optional SKU DDR Capacity: 32 GB/64 GB
Speed Support: 5600/6400/7200 MT/s
Max Capacity Support: 128 GB

*Memory shipment specification is 5600MT/s

Memory Type: DDR5 (SO-DIMM Slot х 2)
Optional SKU DDR Capacity: 32 GB/64 GB
Speed Support: 5600/6400/7200 MT/s
Max Capacity Support: 128 GB

*Memory shipment specification is 5600MT/s