Local LLM - Windows Setup (LM Studio)

Running a local LLM keeps all data private and offline. There are no subscription fees. Hardware and electricity costs apply.

LM Studio is an alternative to Ollama. It uses the same models and the same OpenAI-compatible API. The choice can be changed in settings at any time.

It requires LM Studio and a capable GPU.

Minimum Hardware

To run Elite Dangerous and the LLM on the same machine, a minimum of an NVIDIA RTX 3060 with 24 GB VRAM is required.

Tip: Elite Intel can be pointed at an LM Studio instance running on a separate PC on your network. If a second machine with a capable GPU is available, the game PC carries no inference load in this configuration.

Recommended Model

Model VRAM Required Notes
tulu-3.1-8b-supernova Q4_K_M ~5 GB āœ… Recommended. Fast, accurate, works great for commands and queries.
tulu-3.1-8b-supernova Q8_0 ~8.5 GB Higher quality, if VRAM headroom is available.
qwen3 8B ~8 GB Experimental. Expect occasional missed commands and hallucinations.

Detailed (very detailed) Video tutorial by @DawnTreaderToolsoftheElite

Step 1 - Install LM Studio

Open PowerShell and run:

irm https://lmstudio.ai/install.ps1 | iex

This installs the lms CLI and the LM Studio runtime. Open a new PowerShell window after installation for the changes to take effect.

Verify it worked:

lms --help

Note: If the LM Studio desktop app is already installed, the lms CLI may already be available. Run lms --help before running the install script.

Step 2 - Download the Model

lms get matrixportalx/Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF

or

lms get Tulu-3.1

and choose the matrixportalx/Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF variant (may be listed as Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF).

To list downloaded models:

lms ls

Step 3 - Start the Server

Load the model and start the inference server:

lms load tulu-3.1-8b-supernova --context-length 8192 --gpu max
lms server start

NOTE: The --context-length 8192 flag is required. Without it, the context window may be too small, causing prompt truncation, failures, and hallucinations.

Verify it is running by opening a browser or another PowerShell window and navigating to:

http://localhost:1234/v1/models

You should receive a JSON list of loaded models. The model ID string in that response is what you will enter in Elite Intel's LLM Model field.

To stop the server:

lms server stop

āš ļø Important: The LM Studio server does not survive reboots. Run lms server start again after each restart, or use one of the auto-start options below.

Step 4 - (Optional) Auto-Start on Boot

Two options are available to keep the server running across reboots.

Option A - Desktop App

If the LM Studio desktop app is installed, this is the simplest approach:

  1. Open LM Studio and press Ctrl + , to open Settings.
  2. Check "Run LLM server on login".
  3. Closing the app minimizes it to the system tray and keeps the server running. It restores automatically on next login.

Option B - Task Scheduler (Headless / No GUI)

  1. Press Win + R, type taskschd.msc, press Enter.
  2. Click Create Task in the right panel.
  3. General tab: Name it LM Studio Server. Check "Run with highest privileges".
  4. Triggers tab: Click New ā†’ "At log on" ā†’ OK.
  5. Actions tab: Click New ā†’ "Start a program".
    • Program/script: %USERPROFILE%\.lmstudio\bin\lms.exe
    • Add arguments: server start

To also load the model automatically, create a batch file instead:

@echo off
%USERPROFILE%\.lmstudio\bin\lms.exe daemon up
%USERPROFILE%\.lmstudio\bin\lms.exe load tulu-3.1-8b-supernova --yes --context-length 8192 --gpu max
%USERPROFILE%\.lmstudio\bin\lms.exe server start

Save it as start-lmstudio.bat in a permanent location (e.g. C:\Scripts\) and point the Task Scheduler action at that file.

Step 5 - Configure Elite Intel

Open the Settings tab in Elite Intel:

  • Leave the LLM Key field blank (local LM Studio does not require one).
  • LLM Address: set to http://localhost:1234/v1/chat/completions. If LM Studio is on another machine, replace localhost with that machine's IP.
  • LLM Model: paste in the model ID string from http://localhost:1234/v1/models.
  • Command LLM: set to the same model ID.
  • Query LLM: set to the same model ID.
  • Click Stop then Start on the AI tab to apply changes.

Community šŸ‘‰MatrixšŸ‘ˆ