Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →To run an AI model locally, install a runtime, download model weights it supports, load those weights into your computer’s memory, and start a chat. LM Studio offers a graphical setup; Ollama provides a command-line option; and llama.cpp offers more control. The right choice depends on how you want to work and whether your computer has enough memory for the model.
What you need before you start
A local AI setup has two essential parts: a runtime that loads and runs the model, and the model’s weights, the files containing its learned parameters. The runtime must support the model’s format, and your computer must have enough available memory to load it along with its other parameters. LM Studio identifies GGUF and safetensors as common model formats; llama.cpp’s documented command-line workflow uses GGUF files. LM Studio’s app documentation and the llama.cpp README explain their respective workflows and formats.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
MINISFORUM MS-02 Ultra Workstation Mini PC, Intel Core Ultra 9 285HX (24C/24T, up to 5.5GHz), PCIe... | $1,659.00 | Buy on Amazon |
| 2 |
|
GMKtec EVO-X2 AI Mini PC Ryzen Al Max+ 395 Superchip 128GB LPDDR5X 2TB SSD | $3,649.99 | Buy on Amazon |
Plan to use an internet connection for the initial runtime and model downloads. Afterward, some workflows can run inference offline. Whether that applies to chat, document features, updates, or connected tools depends on the application and how you have configured it.
Choose a setup path
| Option | Best suited to | How you start |
|---|---|---|
| LM Studio | A graphical desktop workflow for finding, loading, and chatting with models | Install the app, download a model in Discover, then load it from the Chat tab. LM Studio app basics |
| Ollama | A command-line workflow, with a local REST API also documented | Install for your operating system, then run ollama run llama3.2. Ollama Quickstart |
| llama.cpp | A more configurable runtime using a local GGUF model file | Install through a package manager, Docker, a prebuilt binary, or a source build; then use llama-cli -m my_model.gguf or run llama-server. llama.cpp README |
These are different setup styles, not a speed or quality ranking. The cited documentation describes workflows and capabilities, not a comparative benchmark.
Recommended Free Tools
#1 Best Overall
- High-Performance AI Processor：The MS-02 Ultra features an Intel Core Ultra 9 285HX (24C/24T, up to 5.5 GHz, 13 TOPS NPU), delivering fast and efficient performance for AI inference, algorithm development, and media workloads. A PCIe x16 expansion slot supports desktop-class GPU upgrades for advanced model training and accelerated computing tasks. It's ideal for creators, engineers, and teams handling intensive parallel workloads.
- 4 × M.2 PCIe 4.0 + 4 × DDR5 SODIMM slots:Four DDR5 SODIMM slots support up to 256 GB of memory, while ECC helps maintain data integrity in mission-critical environments. Four PCIe 4.0 M.2 slots support up to 24 TB of storage, supporting RAID 0/1/5/10, combining high-speed performance with data protection. It allows for the creation of independent scratch disks, media libraries, and project drives, providing high-throughput for production workflows.
- PCIe & USB 4.0 v2: Up to three PCIe slots can be equipped, including a dual-slot x16 GPU. The main slot supports PCIe 5.0, meeting the needs of high-bandwidth creative and computing workloads. USB 4.0 v2 (80Gbps) supports high-bandwidth external storage and displays.
- Ultra-fast Networking: Wi-Fi 7 further enhances wireless performance with next-generation speeds and low-latency stability. Intelligent bandwidth switching optimizes throughput in different network environments, ensuring optimal performance for enterprise or local networks. Dual 25GbE ports (providing up to approximately 3.125 GB/s bandwidth, about 25 times faster than traditional 1GbE), enabling seamless large-scale file transfers and parallel computing. 10GbE and 2.5GbE ports, with support for Intel vPro technology, ensure enterprise-grade remote management and deployment flexibility.
- Server-grade thermal architecture: Utilizing a dedicated CPU/GPU airflow design, equipped with a 6-pipe dual-fan cooler, it maintains stable performance even under sustained loads, delivering up to 140W Turbo power while maintaining a 100W TDP, and operating with noise levels as low as 36 dB. An integrated 350W power supply ensures stable and reliable output for demanding computing tasks and fully loaded extended configurations.
Run your first local chat
LM Studio: use the desktop app
- Install the latest LM Studio app for your computer.
- Open Discover, choose a model, and download it. Check its format and file size before downloading.
- Open the Chat tab and load the downloaded model with the model loader.
- Enter a prompt and send it to start a conversation. Loading allocates memory for the model weights and other parameters.
For LM Studio’s own description of downloading, loading, and chatting, see the app basics documentation.
Ollama: use the command line
- Install Ollama using the official instructions for your operating system.
- Open a terminal and run
ollama run llama3.2. Ollama’s Quickstart uses this command to launch a model; follow the prompts if the model needs to be downloaded first. - Type a prompt in the Ollama session and press Enter to chat.
For other documented management commands, ollama list shows available models, ollama ps shows running models, and ollama stop stops a model. Ollama also documents a local REST API. See the Ollama Quickstart for current installation instructions and usage details.
llama.cpp: load a local GGUF file
- Install llama.cpp using a method in its README, such as a package manager, Docker, a prebuilt binary, or a source build.
- Obtain a GGUF model file that is compatible with the runtime and your available memory.
- In a terminal, run
llama-cli -m my_model.gguf, replacingmy_model.ggufwith the path to your file. - For a server workflow instead of an interactive command-line session, use the documented
llama-serveroption.
See the llama.cpp README for installation options and details on its CPU, accelerator, and hybrid CPU/GPU backends.
Check memory before choosing a model
Model size is a practical constraint: the computer needs to hold the weights and runtime parameters in memory, while leaving room for the operating system and other applications. Ollama’s Quickstart lists Llama 3.2 1B at a 1.3 GB download and Llama 3.2 3B at a 2.0 GB download. These are the listed download sizes, not a guarantee of the total memory needed while running.
Ollama’s Quickstart advises that you should have at least 8 GB of available RAM for 7B models, 16 GB for 13B models, and 32 GB for 33B models. Treat this as Ollama’s guidance, not a universal hardware specification: actual fit depends on the runtime, model format, context, and other applications running at the same time. Ollama Quickstart
Rank #2
- EVOLUTION RYZEN AI MAX+ 395 MINI PC - GMKtec EVO-X2 is the next evolution in AI mini PC Ryzen Strix Halo series. Thanks to AMD Simultaneous Multithreading (SMT) the core-count is effectively doubled, to 32 threads. Ryzen AI Max+ 395 has 64 MB of L3 cache and can boost up to 5.1 GHz, depending on the workload. The Ryzen AI Max+ 395 is currently rated as the "most powerful x86 APU" on the market for AI computing.
- AI NPU with XDNA 2 ARCHITECTURE - Powered by 16 “Zen 5” CPU cores, 50+ peak AI TOPS XDNA 2 NPU and a truly massive integrated GPU driven by 40 AMD RDNA 3.5 CUs, the Ryzen AI MAX+ 395 is a transformative upgrade and delivers a significant performance boost over the competition. The Ryzen AI Max+ 395 excels in consumer AI workloads like the llama.cpp-powered application: LM Studio. Shaping up to be the must-have app for client LLM workloads, LM Studio allows users to locally run the latest language model without any technical knowledge required and unleash their creativity and productivity.
- AMD RADEON 8090S iGPU GAMING PC - The AMD Radeon RX 8060S offers all 40 CUs with up to 2.9 GHz graphics clock and uses the new RDNA 3.5 architecture. The powerful iGPU is positioned between an RTX 4060 and 4070 laptop GPU and therefore enables gaming in FHD at maximum details in most demanding games. The 8060S can also utilize the full 128GB pool, which is perfect for running LLMs such as Deepseek 70B Q8, which runs comfortably on this machine.
- EIGHT CHANNEL LPDDR5X - LPDDR5X is a new ground breaking memory small form factor installed on-board. With blazing speeds up to to 8000MT/s, it runs 1.5x faster than the DDR5 SODIMMs; 90% better performance over DDR5 SODIMMs in video conferencing and photo editing; 30% better performance in productivity apps; 12% better performance in digital content workloads.
- QUAD SCREEN 8K DISPLAY SUPPORT - EVO-X2 AI Mini PC support 4-screen 4K/8K output via HDMI 2.1 (8K@60Hz), DisplayPort 1.4 (4K@60Hz), and dual USB 4 40Gbps Transfer speed (supporting PD3.0/DP1.4/DATA). Ideal for gaming, video editing, and multitasking, it provides expansive and crisp multi-display support.
- Start with a smaller model if you are unsure how much memory is available.
- Compare the model’s download size and format with the runtime’s requirements; download size alone does not establish runtime memory needs.
- Leave memory headroom rather than assuming all installed RAM is available to the model.
Understand offline use and privacy
LM Studio says that, once a model is on the device, its chat, document chat/RAG, and local server can work without an internet connection. Its documentation says chat content and documents remain on the device. Searching for models, downloading models or runtimes, and checking for updates require network access. These statements describe the LM Studio workflow, not every local-model application. LM Studio offline-use documentation
Ollama’s Privacy Policy states: “We do not collect, store, transmit, or have access to your prompts, responses, model interactions, or other content you process locally.” The policy also says limited device and usage metadata may be collected and distinguishes local use from cloud-hosted models, where prompts and responses are processed transiently. This is Ollama’s policy statement, not a guarantee about another runtime or an integration. Ollama Privacy Policy
Local inference does not by itself make every part of a workflow private or offline. If you are handling sensitive material, check the runtime’s current policy, enabled extensions and connected services, whether a server is reachable remotely, and whether the model or other features use a cloud service.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesChoose based on how you want to work
- Choose LM Studio if you prefer a graphical app for finding and loading a model and chatting.
- Choose Ollama if you prefer a terminal command or want to use its documented local API.
- Choose llama.cpp if you want a configurable runtime and are comfortable selecting a compatible GGUF file and working with installation options.
Whichever path you choose, confirm that the runtime supports the model’s format and that the model can fit your available memory before downloading it. Runtime and model catalogs, compatibility, and privacy terms can change; consult each project’s current documentation for the latest details.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
