AI Blocks › Build and ship software
Small-device AI deployer
Give it a trained AI model and a target chip. Get lean code that runs the model on that hardware.
Small-device AI deployer is a free AI skill for the job "Build and ship software". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · BSD 2 Clause licence · Made by MathWorks · Level: expert · Thinking: high
Made by MathWorks. See the original on GitHub; we copied it as it is, with its licence.
What it does
A trained AI model turned into lean code that runs on a small chip, such as a microcontroller. You give the model and name the target hardware, and the skill carries it through to code you can load. It takes models built in MATLAB, or imported from PyTorch, ONNX, TensorFlow and LiteRT, and checks they work. It can shrink them with quantization, test them inside a Simulink system model, and generate C, C++ or CUDA code for Cortex-M, A or R chips, x86 and GPUs. This needs a paid MathWorks setup: a recent MATLAB with the deep learning, coder and embedded toolboxes, plus the MATLAB agent connection for your AI.
You give it
- The trained model
- The target chip and its memory limits
- MATLAB with the needed toolboxes
- Accuracy you can accept after shrinking (optional)
You get
- The model checked and imported
- A compressed version sized for the chip
- Generated C, C++ or CUDA code
- A system simulation of the model (optional)
Use it when
- You have a trained network and must run it on a microcontroller.
- Your model is too large for the memory on your device.
- You want to test a model inside a larger system simulation first.
- You need C code generated from a PyTorch model.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run MATLAB with Embedded Coder.