CrankGPT
A hand-cranked, fully offline voice assistant.
By Squeez Labs
- GitHub stars
- 62 (GitHub)
Build sheet
- Board
- Raspberry Pi 5 8GB
- Other parts
- KEYESTUDIO ReSpeaker 2-Mic HAT, 20W hand-crank generator, custom supercapacitor power board
- AI model or service
- Moonshine ASR, Silero VAD, llama.cpp (LFM2.5 350M/1.2B or Gemma 3 1B), Piper TTS
- Runs offline?
- Yes, per the source
- Difficulty
- Advanced
Overview
Squeez Labs runs Moonshine ASR, a small local LLM via llama.cpp (LFM2 / Gemma 3 1B), and Piper TTS on a Raspberry Pi 5 powered only by a 20W hand-crank generator plus a custom supercapacitor board—no battery, no cloud.
How it works
Moonshine ASR -> small llama.cpp model (Gemma 3 1B or LFM2) -> Piper TTS on a Pi 5, powered only by a hand-crank generator and a supercapacitor board.
Hardware
- Raspberry Pi 5 8GB
- KEYESTUDIO ReSpeaker 2-Mic HAT
- 20W hand-crank generator
- custom supercapacitor power board
Like this build? Get the weekly builds
Get the best new AI hardware builds every Thursday. One email a week, free.
Built something? Submit a build
This page is a short summary written by HardwareLab. Build steps, code and photos belong to the creator; follow “View original” for the full project.
Featured? Add this badge
Built this? Paste the badge into your README. It links back to this page.
Markdown
[](https://hardwarelab.dev/projects/crankgpt/?utm_source=badge&utm_medium=referral&utm_campaign=launch_2w)HTML
<a href="https://hardwarelab.dev/projects/crankgpt/?utm_source=badge&utm_medium=referral&utm_campaign=launch_2w"><img src="https://hardwarelab.dev/badge.svg" alt="Featured on HardwareLab" width="188" height="20"></a>Related builds
All 43 in Voice Assistants & CompanionsPrivate offline Pi voice assistant
Whisper + a small LLM + Piper, all on a Pi.
- Raspberry Pi
- mic
- speaker
Beginner
AImy
Offline vision-enabled voice assistant on a Pi 5 with an LLM accelerator.
- Raspberry Pi 5
- LLM-8850 accelerator
- camera
Advanced23
