Overview

IPEX-LLM is an LLM acceleration library for Intel GPU (e.g., local PC with iGPU, discrete GPU such as Arc, Flex and Max), NPU and CPU This Image is meant to be a one way stop for AI enthusiasts with an Intel iGPU or dGPU to run models using Ollama. Make sure to have reBar enabled

Requirements

Unraid 7+

Runtime arguments

Network: bridge
Shell: bash
Privileged: false

Template configuration

ModelsPathrw

Target: /root/.ollama
Value: /mnt/user/Models/ollama

GPUDevice

Target: /dev/dri
Value: /dev/dri

OLLAMA_HOSTVariable

Allow connections from the Value

Target: OLLAMA_HOST:
Default: 0.0.0.0
Value: 0.0.0.0

ONEAPI_DEVICE_SELECTORVariable

Value: level_zero:0

IPEX_LLM_NUM_CTXVariable

Value: 16384

Ollama portPorttcp

Target: 11434
Value: 11434

OLLAMA_INTEL_GPUVariable

Value: 1

OLLAMA_NUM_GPUVariable

Value: 999

ZES_ENABLE_SYSMANVariable

Value: 1

OLLAMA_DEBUGVariable

Enable Ollama debug logs

Value: 1

Links

Template Support GHCR

Details

Repository

ghcr.io/justjoseorg/ollama-intel-gpu:main

Registry

https://ghcr.io/justjoseorg/ollama-intel-gpu

Last Updated2026-05-31

First Seen2025-04-20

Run Intel-IPEX-LLM-Ollama on Unraid.

Intel-IPEX-LLM-Ollama is listed in Community Apps for Unraid OS. Explore Unraid to build a flexible home server, NAS, or homelab.

Explore Unraid OS

Intel-IPEX-LLM-Ollama