You could build a local AI box yourself. PHNTM One is the version where the tuning, the burn-in and the privacy plumbing are done, and nothing is locked. It's your machine, and it's meant to be opened up.
What's yours
- Debian 13 on a Raspberry Pi 5 with 8 GB, with Ollama bound to localhost and run with
--offline.
- SSH, documented and never removed. The username and password you choose during setup are the device's login.
- Readable source on the device under the PHNTM One Source License, so you can check what it does.
- A local HTTP API on port 8800, the same one the touchscreen uses, gated by your pairing token and reachable only from your LAN.
- The full bill of materials, every part and every dollar. See the BOM →
Why not just run Ollama yourself?
You can, and the buyer's guide compares both honestly. PHNTM One adds the parts that take weeks: a RAM budget that keeps the model resident, local memory and document search, tap-to-talk voice, phone pairing with no relay, a firewall that stops the model server reaching the internet, self-healing services, and updates that roll back if a self-test fails.
Questions
Can I change the model?
It's your Debian machine, so you can experiment. The shipped configuration is tuned for Gemma 3 4B within the 8 GB budget; bigger models don't fit.
Is the code open source?
It ships as readable source on the device under the PHNTM One Source License. You can read it and verify it, but it isn't an open-source license. See /credits for everything open it's built on.