🧠 Running MESH-AI on a Raspberry Pi

[mr-tbot]

🧠 Running MESH-AI on a Raspberry Pi (Pi 4 or 5 Recommended)

⚠️ Raspberry Pi 4 or 5 is recommended — older models may not handle AI workloads effectively.

💡You can also run MESH-AI inside of a docker container - ARM7 releases for Raspberry Pi are available on dockerhub - mrtbot/mesh-ai:latest - https://hub.docker.com/r/mrtbot/mesh-ai

---

📦 Step 1: Update Your System

sudo apt update && sudo apt upgrade -y

---

🧠 Step 2: Install Ollama and Download a Language Model

MESH-AI uses Ollama to run local LLMs.

Install Ollama:

sudo apt install curl -y
curl -fsSL https://ollama.com/install.sh | sh

Verify the Installation:

ollama --version

Run the Model Once to Download & Initialize:

ollama run hf.co/mradermacher/Llama-3.2-1B-Instruct-Uncensored-GGUF:Q4_K_M

Once it loads successfully, press Ctrl+C to exit.

💡 This primes the model so MESH-AI can use it later.

---

🐍 Step 3: Install Python 3.13 from Source

The default Python (3.11) is outdated — install Python 3.13 for best compatibility.

Install Required Build Tools:

sudo apt update && sudo apt install -y \
  build-essential libssl-dev zlib1g-dev libbz2-dev \
  libreadline-dev libsqlite3-dev wget curl llvm \
  libncursesw5-dev xz-utils tk-dev libxml2-dev \
  libxmlsec1-dev libffi-dev liblzma-dev git

Download and Build Python 3.13.2:

cd /usr/src
sudo wget https://www.python.org/ftp/python/3.13.2/Python-3.13.2.tgz
sudo tar xzf Python-3.13.2.tgz
cd Python-3.13.2
sudo ./configure --enable-optimizations
sudo make -j1
sudo make altinstall

⚠️ Use make altinstall to avoid overwriting system Python.

---

🔄 Step 4: Set Python 3.13 as Default

/usr/local/bin/python3.13 --version
sudo update-alternatives --install /usr/bin/python3 python3 /usr/local/bin/python3.13 1
sudo update-alternatives --set python3 /usr/local/bin/python3.13
python3 --version

---

📥 Step 5: Install pip for Python 3.13

curl -sS https://bootstrap.pypa.io/get-pip.py | python3

If that fails:

curl -sS https://bootstrap.pypa.io/get-pip.py -o get-pip.py
sudo /usr/local/bin/python3.13 get-pip.py

Optional: Fix Missing pip Command

sudo ln -s /usr/local/bin/pip3.13 /usr/bin/pip3
sudo ln -s /usr/local/bin/pip3.13 /usr/bin/pip

---

🧼 Optional: Remove Older Python Versions

Only if everything is working:

sudo apt remove --purge python3.11 python3.9 python3.7 -y
sudo apt autoremove -y

---

🚀 Step 6: Clone and Set Up MESH-AI

git clone https://github.com/mr-tbot/mesh-ai.git
cd mesh-ai

Create and Activate Virtual Environment:

python3 -m venv mesh-ai
source mesh-ai/bin/activate

Install Dependencies:

pip install --upgrade pip
pip install -r requirements.txt

Configure Files:

Edit the following as needed:

• config.json
• commands_config.json
• motd.json

If using USB / Serial - Add user to dialout group

ls /dev/ttyUSB* /dev/ttyACM*

Identify your connected device - typically /dev/ttyUSB0

Check Permissions

ls -l /dev/ttyUSB0

You'll see something like this:

crw-rw---- 1 root dialout 188, 0 Jul 28 21:14 /dev/ttyUSB0

Then simply run the following while running as your intended user. Not root:

sudo usermod -aG dialout $USER
newgrp dialout  # or reboot

---

▶️ Step 7: Run the Bot Manually (Test)

python mesh-ai.py

Then open http://localhost:5000/dashboard in your browser.

---

🔁 Step 8: Auto-Start MESH-AI at Boot

Create Startup Script

Create a file named start-mesh-ai.sh in your project folder:

cd ~/mesh-ai
nano start-mesh-ai.sh

Paste this in:

#!/bin/bash
cd /home/pi/mesh-ai
source mesh-ai/bin/activate
python mesh-ai.py

Replace /home/pi/mesh-ai with your actual directory if different.

Make it executable:

chmod +x start-mesh-ai.sh

---

Create systemd Service

sudo nano /etc/systemd/system/mesh-ai.service

Paste this:

[Unit]
Description=MESH-AI Bot
After=network.target

[Service]
Type=simple
User=pi
WorkingDirectory=/home/pi/mesh-ai
ExecStart=/home/pi/mesh-ai/start-mesh-ai.sh
Restart=on-failure
Environment=PYTHONUNBUFFERED=1

[Install]
WantedBy=multi-user.target

Make sure to adjust User= and paths to match your setup.

---

Enable and Start the Service

sudo systemctl daemon-reexec
sudo systemctl daemon-reload
sudo systemctl enable mesh-ai.service
sudo systemctl start mesh-ai.service

---

Check Service Status

sudo systemctl status mesh-ai.service

To view live logs:

journalctl -u mesh-ai.service -f

---

✅ Done!

MESH-AI is now running on your Raspberry Pi, auto-starts at boot, and is ready to respond to messages with the help of your local Ollama Server.

[mr-tbot]

You can also now use docker to run this on RPI (and other ARM based platforms) - see the READ ME for more details!

[mr-tbot]

These instructions have been slightly updated based on personal experience. Hopefully this helps!