Deploy in the cloud or at the edge
OctoML produces fully tuned models — pre-packaged, pre-configured, and ready for deployment with a few simple lines of code.
Get cross-device portability
Run your models automatically, with a single API and consistent packaging story, on a variety of hardware:
CPUs, GPUs, accelerators, IoT devices, and smartphones.
Choose from flexible packaging options
We provide a variety deployment packaging deployments that snap into your application with ease, including: Python wheel, Docker, shared library with C API, serverless cloud tarball, and more.