Take your machine learning models into production
DeepMux is a platform to deploy machine learning models into production. Our system allows you to use only as much GPU time as you really need. Just upload your model and get predictions, zero tweaking required.
Key features
Our novel approach to model deployment is what makes us different
Flexible pricing
Billing is performed on per inference basis, we only take into account the execution time and model size.
Automatic scaling
Zero worries about scaling, as your business grows, we will automatically provide you with more resources to keep up with your growth.
Serverless
Move beyond physical machine limitations. Let your developers focus on core product and cut your time and costs on managing and operating servers.
Run from anywhere
Edge or backend, frontend or mobile. We provide you with runtime SDK for most common platforms and languages.
GPU support
Inference fast and efficiently. Get an ability to run heavy models requiring high parallelism.
CV ready
Host your production Computer Vision models effortlessly.
Client SDKs
.
Supported frameworks
Models in following formats can be uploaded to DeepMux
PyTorch
ONNX
TensorFlow
coming soon
coming soon