Skip to content

Install Ollama Local LLM Linux

Jump to bottom
alex edited this page May 6, 2026 · 10 revisions

Local LLM - Linux Setup (Ollama)

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

It requires Ollama 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 12 GB VRAM is required. Performance headroom is limited at this specification.

Tip: Elite Intel can be pointed at an Ollama 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 ✅ Recommended. Reliable for commands and queries.
qwen3 8B ~8 GB Experimental. Expect occasional missed commands and hallucinations.

Note: For the fastest local inference, consider LM Studio with matrixportalx/tulu-3.1-8b-supernova. In testing, it is noticeably faster than Ollama on the same hardware with the same model.

Step 1 - Install Ollama

curl -fsSL https://ollama.com/install.sh | sh

Ollama installs as a systemd service and starts automatically.

Step 2 - Pull a Recommended Model

ollama pull hf.co/matrixportalx/Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF

Or experimental alternatives:

ollama pull qwen3:8b

Step 3 - (Optional) Tune the Ollama Service

Ollama works without tuning. The following configuration improves VRAM management when running alongside Elite Dangerous.

sudo nano /etc/systemd/system/ollama.service.d/override.conf

Paste this in:

[Service]
Environment="OLLAMA_MAX_VRAM=14000000000"
Environment="OLLAMA_DEBUG=0"
Environment="OLLAMA_NUM_PARALLEL=3"
Environment="OLLAMA_MAX_LOADED_MODELS=1"
Environment="OLLAMA_FLASH_ATTENTION=1"
Environment="OLLAMA_KEEP_ALIVE=-1"
Nice=10
IOSchedulingClass=best-effort
IOSchedulingPriority=5

Then reload and restart:

sudo systemctl daemon-reload
sudo systemctl restart ollama.service

What these settings do

OLLAMA_MAX_VRAM: Hard cap on VRAM Ollama can use, in bytes. 14000000000 = 14 GB. Leaves the remainder for Elite Dangerous. Adjust based on your GPU and game requirements.

OLLAMA_NUM_PARALLEL: Number of requests Ollama handles simultaneously. Elite Intel makes async calls, so setting this too low causes failures. 3 covers the typical command and query overlap without over-allocating.

OLLAMA_MAX_LOADED_MODELS: Keeps only one model in VRAM at a time.

OLLAMA_FLASH_ATTENTION: Enables Flash Attention, which reduces memory bandwidth usage during inference. Generally faster, especially for repeated requests.

OLLAMA_KEEP_ALIVE=-1: Keeps the model loaded in VRAM indefinitely. Without this, Ollama may unload the model after a period of inactivity, incurring a reload penalty on the next request.

Step 4 - Configure Elite Intel

Open the Settings tab in Elite Intel:

  • Leave the LLM Key field blank (local Ollama does not require one).
  • LLM Address defaults to http://localhost:11434/api/chat. If Ollama is on another machine, replace localhost with that machine's IP.
  • Command LLM: set to hf.co/matrixportalx/Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF:latest (or the name shown by ollama ls).
  • Query LLM: set to hf.co/matrixportalx/Tulu-3.1-8B-SuperNova-Q4_K_M-GGUF:latest (or the name shown by ollama ls).
  • Click Stop then Start on the AI tab to apply changes.

Community 👉Matrix👈

Clone this wiki locally