Ship bare-metal AI to your firmware in seconds

Compile ONNX and PyTorch models into lean, dependency-free C++ for microcontrollers and the edge.

qovox_model.h

// generated by Qovox · no libc++, no heap
#pragma once
#include <stdint.h>

namespace qovox {
  constexpr size_t ARENA = 9216;
  void infer(const int8_t* in, int8_t* out);
}

arena 9 KBheap 0 Bdeps none

⚡Zero-dependency C++ export for Cortex-M & ESP32
🔒Enterprise-grade model encryption & privacy
🚀Up to 4.2× smaller footprint with INT8 quantization

Built for firmware & embedded teamsSTM32ESP32ARM Cortex-MRISC-V

← Back to site

Create your Qovox account

Compress ONNX and PyTorch models into C++ for the edge.

Or continue with email

Full name

Username

3–20 characters: a–z, 0–9 and underscore.

Work email

Password

Confirm password

8+ characters
Uppercase letter
Lowercase letter
Number
Special character

What do you do?

How did you find us?

By creating an account you agree to our Terms and Privacy Policy.

Work email

Password

Already have an account?

Finish your profile

Pick a username and tell us what you build. This takes ten seconds.

Full name

Username

3–20 characters: a–z, 0–9 and underscore.

What do you do?

How did you find us?

Check your email

We sent a 5-digit code to . It expires in 10 minutes.