Raspberry Pi Project Finder

Find a Raspberry Pi project that fits your board, experience, free time and the parts you already have. Compare what you’ll need and save a checklist for your next build.

Find a project

Parts you own

Costs are rough USD estimates for add-ons. Check each project for the board, power supply, storage and other equipment outside these filters.

9 matching projects

What can you self-host on a Raspberry Pi?

DNS filtering, automation and file sharing
A Zero 2 W can serve a small Pi-hole setup. A Pi 4 or Pi 5 gives you more room for an automation hub, a file share or direct-play media. Start with one service, plan for updates and keep backups off the Pi.
Local models and API-backed assistants
On an 8 GB Pi 5, try a quantized 1–3B text model with short context and measure how fast it responds. Large models, long context and multiple users can strain memory and CPU. An API-backed assistant can run its app on the Pi while inference happens on a hosted model.

If you want an always-available OpenClaw assistant without maintaining a Pi, NitroClaw offers managed hosting.

Explore NitroClaw hosting

Power supplies by board

Raspberry Pi 5
27 W USB-C PD (5 V / 5 A) recommended
Raspberry Pi 4
15 W USB-C (5 V / 3 A)
Raspberry Pi 3
12.5 W Micro USB (5 V / 2.5 A)
Raspberry Pi Zero 2 W
12.5 W Micro USB (5 V / 2.5 A)
Raspberry Pi Pico
USB power with a data-capable cable for setup
Raspberry Pi power supply guide

Frequently asked questions

What Raspberry Pi project should I build first?

A desk weather station is a practical starting point if you have a compatible sensor and display. If you only have a Pi computer with storage and network access, try a DNS filter on one test device.

Can a Raspberry Pi Pico run the same projects as a Pi 4?

The Pico is a microcontroller for tasks such as reading sensors and controlling displays. A Pi 4 runs a full Linux operating system and can host a file share, media server or automation hub.

Can I run a local AI model on a Raspberry Pi 5?

A Pi 5 with 8 GB of RAM is a starting point for testing small quantized text models with short prompts. Model size, context length and runtime affect memory use and speed. Benchmark before depending on it.

Where is my shortlist saved?

The finder saves your shortlist in this browser. You can print it for reference. Clearing site data or switching browsers can remove access to the saved list.