brenpoly

brenpoly / be-more-agent

Public

Local AI Agent running on Raspberry Pi

youtu.bel5ggH-YhuAw
284
42
100% credibility
Found Feb 11, 2026 at 40 stars 7x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

This project builds a fully local, voice-activated AI assistant on Raspberry Pi that listens for a wake word, understands speech, responds with a synthesized voice, shows reactive face animations, and can search the web or describe images from a camera.

How It Works

1
📱 Discover the cute robot

You find a fun demo video of a little animated assistant that chats and reacts on a Raspberry Pi.

2
🛒 Gather simple hardware

Pick up a Raspberry Pi, USB microphone, speaker, screen, and optional camera to bring it to life.

3
🛠️ Run easy setup

Follow the one-click setup script that prepares everything automatically so you're ready to go.

4
🎨 Customize your character

Swap in your own face pictures and fun sound clips to give it a unique personality.

5
🗣️ Wake it up and chat

Say the wake word like 'Hey Jarvis', speak your question, and watch it listen, think, search, or describe what it sees.

🤖 Your private AI buddy lives!

Enjoy endless offline conversations with your helpful, animated robot companion that feels truly alive.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 40 to 284 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is be-more-agent?

Be-more-agent turns a Raspberry Pi into a fully local AI agent that listens for a wake word, transcribes speech with Whisper.cpp, reasons with Ollama LLMs like Gemma, and responds via Piper TTS—all offline. It handles vision tasks with Moondream, pulls real-time info via DuckDuckGo search for agentic RAG local workflows, and shows reactive face animations on screen. Developers get a customizable local agent AI framework in Python, swap faces and sounds to build unique characters without cloud dependencies.

Why is it gaining traction?

It stands out as a local agent LLM that runs smoothly on Pi 4/5 hardware, dodging API costs and privacy issues of cloud agents, with easy setup via a single script that pulls models and voices. The hook is instant customization—drop in PNG sequences and WAV files for personality—and features like PTT mode, adaptive recording, and camera integration make it feel polished for local agentic coding experiments. Early adopters dig the zero-config path to a be more agentic desktop companion versus bloated alternatives.

Who should use this?

Raspberry Pi makers building local agent services for home automation or exhibits. Hobbyists in Australia tweaking local agent finder bots for offline queries. Tinkerers wanting a local GitHub Copilot alternative that describes camera feeds or searches news without internet handoffs.

Verdict

With 19 stars and 1.0% credibility score, it's an early-stage local agent GitHub gem—solid docs and demo video lower the entry barrier, but expect tweaks for production. Grab it if you're prototyping local GitHub runner-style agents on Pi; skip for battle-tested scale.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.