Included in Every Setup
Open WebUI
ChatGPT-style AI, on your own hardware
Open WebUI is a browser-based interface for local AI — it looks and works like ChatGPT, but runs entirely on your Mac. Multiple users can connect, conversations are saved locally, and you can switch between any installed models. Included in every setup we do.
Book a consultation →What it is
The familiar experience, without the cloud
Open WebUI gives you a chat interface that anyone can use immediately — it feels just like ChatGPT. Conversations, model switching, history, multimodal inputs — all the features you expect, running on hardware you own.
Unlike ChatGPT, your conversations are stored locally on your Mac. There's no data harvesting, no profile building, no advertising model. What you type stays in your house.
Open WebUI is the baseline interface in every setup we do. Whether you're also running OpenClaw, AnythingLLM, or Cline, Open WebUI is always there as the familiar fallback — open a browser tab and start chatting with any installed model.
How it works
Browser chat, local inference
Open WebUI runs as a local web server on your Mac. You open it in any browser — on your Mac, your phone, or any device on your network via Tailscale. The interface talks to Ollama running locally; all inference stays on your hardware.
We set up user accounts so multiple household or team members can each have their own conversation history. Each user's chats are private to them. Model selection is available per-conversation — switch between any installed model from the dropdown.
Who it's for
Everyone — it's the baseline for all setups
- ✓Families who want a private, ad-free AI for homework, recipes, and research
- ✓Teams sharing a Mac Mini who each want their own AI conversation history
- ✓Anyone transitioning from ChatGPT who wants the same experience locally
- ✓Users who want a simple, no-configuration chat interface
- ✓Anyone who needs a quick answer without switching to their primary framework
Included in every setup
Open WebUI is installed in every setup we do — OpenClaw, AnythingLLM, Cline, all of them. It's the browser interface from which you access all installed models. You don't need to choose it as your primary framework; it's already there.
Full stack
What gets installed
| Layer | Component | Purpose |
|---|---|---|
| AI Engine | Ollama (MLX backend) | Runs models on Apple Silicon |
| Chat Interface | Open WebUI | Browser-based chat with model switching and history |
| Networking | Tailscale | Secure access from anywhere, any device |
| Multi-user | User accounts | Multiple family or team members, each with their own history |
| Security | Hardened config | Loopback binding, authenticated access |
| Monitoring | Health watchdog | Auto-restart on crash |
Security
Private conversations, local storage
All conversations are stored in a local database on your Mac — not in the cloud, not on Open WebUI's servers. We configure authentication so only authorized users can access the interface. Access from outside your network goes through Tailscale, not an open port.
Recommended models
Works with any model
Open WebUI supports all models installed through Ollama — switch between them in the interface
Great everyday model — fast, capable, fits on all Mac Minis
Frontier-class for larger Macs — GPT-4 equivalent quality
Open WebUI is included in every setup.
Book a consultation. We'll configure your complete AI stack — Ollama, Open WebUI, and your primary framework — and hand it back ready to use.
Book a Consultation