Your Free ML Deployment Tool
Get a trained model running in your app, in your environment, in a few easy steps.
The OctoML CLI: Your Path to Production
You built and trained the perfect ML model for your application, now it’s time to push to production. You don’t need to spend months learning a new set of tools and techniques – you need a bridge to get from model development to deployment as fast as possible.
With the OctoML Command Line Interface (CLI), you can rapidly deploy ML using:
Your choice of models
Your environment
Your application
Your cloud or hardware targets
The OctoML Platform optimizes and containerizes ML models into a flexible, CI/CD friendly deployment package that can be repackaged and redeployed in a few clicks when the model or hardware target changes.
When you’re ready, use OctoML to accelerate the model with various engines such as Apache TVM, ONNX-Runtime, TensorRT, and more to get the maximum performance for the hardware you choose.
Download the free OctoML CLI
Trained ML models become agile, portable, reliable software functions that easily integrate with your workflows.
Learn how with Transparent AI
Adding a model to the application
Exposing the model over an API
Wrapping that API with pre & post processing code
Deploying the model in the app for development
Deploying the model with the app for production in the cloud
Monitoring the model as it runs in production
Accelerating the model for best performance
Built on Open Source technologies, and enabled by OctoML, Transparent AI empowers developers to add intelligence to their applications. Get started with Transparent AI.
OctoML Customers and Partners
Automate Model Deployment to Any Hardware
The CLI cleans up the trained model artifact –model math expressed in Python code– and streamlines it into a portable, intelligent function to be used your way.