Let’s make your AI
tiny and fast.

Try out our inference engine for Arm Cortex-M. It is the fastest and smallest in the world. MobileNetV2 runs with 40% lower latency and uses 49% less RAM. Model accuracy does not change. No binarization, no additional quantization, no pruning.

Here's how it works

Upload your .tflite model.

We compile and run it on Arm Cortex-M4 (STM32L4R9) and Arm Cortex-M7 (STM32H7B3) microcontrollers.

We show the speedups and memory savings with our inference engine compared to TensorFlow Lite for Microcontrollers.

Model requirements

TensorFlow Lite Flatbuffer .tflite file.
INT8 quantized, float is not supported.
Model size smaller than 2MB.
Runs with TensorFlow Lite for Microcontrollers.

Contact us and use our inference engine in your product

Get started

Let’s make your AI
tiny and fast.

Drag .tflite file here

filename

We will not reuse, reverse engineer, or steal your model, and will not sell or provide your data to third parties.
For full details, see our privacy policy.

Here's how it works

Model requirements

Contact us and use our inference engine in your product

Let’s make your AItiny and fast.

Drag .tflite file here

Here are some sample models:

filename

We will not reuse, reverse engineer, or steal your model, and will not sell or provide your data to third parties. For full details, see our privacy policy.

Here's how it works

Model requirements

Contact us and use our inference engine in your product

Let’s make your AI
tiny and fast.

We will not reuse, reverse engineer, or steal your model, and will not sell or provide your data to third parties.
For full details, see our privacy policy.