deputyOS on Raspberry Pi 4
Run deputyOS on a Raspberry Pi 4. Works on the lean tier with cloud model providers; local inference is out of reach.
- Setup time
- ~20 min
- Difficulty
- Easy
- Cost
- ~$45 one-off
- Acceleration
- native arm64
The Pi 4 is supported but constrained. It runs the lean tier only: ClamAV scans on demand rather than resident, and there is no local LLM or voice.
If you are buying hardware, buy a Pi 5. If you have a Pi 4 in a drawer, this works.
What you need
- Raspberry Pi 4, 4 GB or 8 GB
- microSD card (32 GB+) or a USB 3.0 SSD, which is noticeably better
- Official USB-C power supply
- Ethernet for setup
Flash and boot
Select rpi4 in the picker and write the .img.xz exactly as for the
Pi 5. Boot takes longer — allow five minutes before the wizard responds — then open
http://deputyos.local.
What you give up on lean
- No local model. Bring an API key for Anthropic, OpenAI, OpenRouter or Google AI Studio.
- ClamAV runs on demand, not resident, so file scanning adds latency when it fires.
- No voice pipeline.
- Heavy recall over large document sets will be slow.
Everything else — channels, memory, tools, scheduling, tunnel, backup — works normally.
Thermals
The Pi 4 throttles under sustained load without cooling, and a throttled Pi looks like a slow assistant rather than a hot one. A heatsink case or a small fan is worth the few pounds.