Guides · Updated 2026-10-17

How to install Ollama on Linux (Ubuntu, Debian, Fedora, Arch)

One-line install, NVIDIA and AMD GPU setup, running Ollama as a systemd service, and exposing it on your network safely.

Not sure which model to pull? Use the GPU model picker: enter your VRAM and RAM and get a list of Ollama models that fit.

On Linux, Ollama installs with a single command and runs as a systemd service, which makes it ideal for a home server or headless GPU box.

Step 1: Install

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

The script installs the binary, creates an ollama system user, and sets up and starts a systemd service. Check it worked:

ollama -v
systemctl status ollama

Prefer to read scripts before piping them to a shell? Download install.sh first, or use the manual tarball install described in the official Linux docs.

Step 2: GPU drivers

  • NVIDIA: install the proprietary driver (550 or newer; older Maxwell/Pascal cards need 570+) and check with nvidia-smi. Compute capability 5.0 or newer is supported.
  • AMD: Ollama’s ROCm build needs ROCm v7. Supported Radeon cards include the RX 9070/9060 series, RX 7900/7800/7700/7600 series and RX 6800/6900 series. Install with AMD’s amdgpu-install tool. Other cards may work through Vulkan; you may need to add the ollama user to the render group.
  • Intel Arc: works through Vulkan. Install Intel’s GPU drivers and Mesa Vulkan packages for your distro.

More on non-NVIDIA cards in our AMD, Intel and mini PC guide.

Step 3: Run a model

ollama run qwen3.5:9b
ollama ps   # PROCESSOR should read "100% GPU"

Configure the service

Settings are environment variables on the service. Run sudo systemctl edit ollama and add, for example:

[Service]
Environment="OLLAMA_CONTEXT_LENGTH=8192"
Environment="OLLAMA_MODELS=/data/ollama"

Then sudo systemctl daemon-reload && sudo systemctl restart ollama. Logs: journalctl -e -u ollama.

Reach it from other machines

By default Ollama only listens on localhost. Setting OLLAMA_HOST=0.0.0.0:11434 exposes it to your network. Ollama has no login of its own, so only do this on a trusted network, or put it behind a reverse proxy with authentication or a VPN such as Tailscale. Never port-forward 11434 to the internet.

Update or uninstall

Update by re-running the install script. To pin a version: curl -fsSL https://ollama.com/install.sh | OLLAMA_VERSION=x.y.z sh.

Get the Local AI weekly

New models worth running, what fits which GPU, and setup fixes. One short email a week, free. Unsubscribe any time.

Lila Forge Kits

Local-LLM App Starter

A tested starter for building your own app on Ollama: chat, RAG and a GPU-fit checker. Skip the setup hours.

kits.lilaforge.com →
VibeEngine

Best local LLM tools, ranked

Ollama, LM Studio, Open WebUI and llama.cpp scored from public evidence, with top picks.

vibeengine.lilaforge.com →