The future of automated AI compression.

ace polygon img

Inference-ready model compression, automated with just one line of code

ace hero image
CLIKA’s ACE(Auto Compression Engine) is the fully-automated solution you need to easily put your winning AI model into production.
It fits perfectly into any AI pipeline and makes deployment onto any device easy and convenient.
ace gif image

Quantization has never been this easy.

Import your PyTorch model and let our compression engine do all the hard work for you. It’s that simple.
Tailor your compression to the specific device or target hardware where you want to deploy your model.
Easy as 1, 2, 3.

1. Import your model

2. Customize your setup

3. Run the compression

Customize your settings
with just a few lines of code.

ace polygon img
ace gif image

Set the target framework

Choose the right format for your hardware needs: ONNXRuntime, TensorRT, OpenVino or TFLite.

Set the number of model bits

Set the number of bits (16, 8, 4 or 2) in which you want to quantize your model.

Set the quantization options

Choose between different quantization options to achieve the best results for your use case.

Set the training mode

Customize settings for training, including multi-GPU support for even faster compression.

Wish your AI compression and compiling jobs
would magically just work?​​