NoLlama

Local LLM server for the full Intel stack. NPU, ARC iGPU, ARC discrete, CPU. OpenAI + Ollama APIs. One server, every Intel device.

No NVIDIA required. No Ollama install. No llama.cpp. No problem.

It detects your hardware, picks the best device, and speaks both OpenAI and Ollama APIs — so any client that talks to either just works. It drives coding agents too: OpenCode, Goose and VS Code Copilot Chat run against it with local tool-calling on your Intel GPU or CPU.

Quick start

No git? Grab the latest release ZIP, unzip, then double-click install-windows.bat (Windows) or run ./install-linux.sh. Both check for PowerShell 7 and Python 3.10+ and offer to install what’s missing.

Already have those, or cloned the repo:

.\install.ps1     # detects hardware, pick a model, writes start.ps1
.\start.ps1       # then open http://localhost:8000

Needs Intel hardware, Python 3.10+, PowerShell 7+. Windows and Linux both work. Details in docs/DEVICES.md.

Use it from anything

Point any OpenAI client at…