Uploaded January 2026 | Updated September 2026, 2 weeks ago

---

Let's dive into the Raspberry Pi AI Hat+ 2 with 40 TOPS. I'll walk you through the initial setup process, then show you how to run a Large Language Model (LLM) on this new board. I'll be using my own project, Pillama, and Open WebUI to demonstrate. Finally, I'll reveal the performance numbers, showcasing the tokens per second generated by the CPU alone versus the AI HAT+ 2 board, so you can see the difference!
---
Pillama: github.com/garyexplains/examples/blob/master/pillama.py
AI HAT+ 2: raspberrypi.com/documentation/accessories/ai-hat-plus.html
AI Software: raspberrypi.com/documentation/computers/ai.html
---
⭐ Please support my channel on Patreon!
Get early access to videos, members-only content, behind-the-scenes updates, and join the Gary Explains Discord! Join here 👉 patreon.com/GaryExplains 🙌

Twitter: twitter.com/garyexplains
Instagram: instagram.com/garyexplains

#garyexplains