Deploy, optimize and orchestrate ML models on

Deploy, optimize and orchestrate ML models on

Unlock the Power of Edge Intelligence

Convert models for multiple runtimes in a single step

Bring your existing TensorFlow, CoreML and PyTorch models or add new ones straight from the training script

Choose your target runtimes (e.g. TF Lite, TensorRT, CoreML, TVM) and let Makinas handle the conversion for each of them

Map models to devices, platforms, apps and users

Once the model is converted and optimized for the target, it can be directly pulled from the device or application, no need to wait for software updates

Makinas works as a content delivery network for your models, to easily serve them across your device ecosystem

Share your work with the community or quickly prototype using public models

Makinas is more than a tool, it is an evergrowing collection of ready-to-use models and tutorials on how to deploy them on the edge

Of course, your models are secure if you prefer to keep them private