What is Model Zoo?¶

Model Zoo is a machine learning model deployment platform with a focus on ease-of-use. Deploy your model to an HTTP endpoint with a single line of code:

from modelzoo.tensorflow import deploy, predict

# Train or load your TensorFlow here.
model = train_model()

# Deploy with one function.
model_name = deploy(model)

# Make predictions from Python.
predictions = predict(model_name, image="test.jpg")


  • ⏱️ Single function deployment, no need to write any code or learn new technologies!

  • 👁️ Realtime monitoring of model features and predictions.

  • 💵 Autoscaling down to zero during periods of low activity to save costs, and up to accomodate bursts of demand.

  • 📝 Auto-generated documentation of the model inputs and outputs.

  • 🖥️ Web interface for testing and sharing models.

  • 🐍 Python client library for making predictions.

And many more to come!

Get Started¶

  1. Take a look at an example model (30 seconds ⚡)

  2. Deploy your first model with a quickstart (5 minutes 👨‍💻 👩‍💻)


Model Zoo is currently free to use for deploying up to three models. Each model in the free tier will automatically be scaled down after 20 minutes of inactivity.

The private beta version of Model Zoo offers unlimited model uploads, in addition to the ability to influence our development roadmap, a direct Slack/email support line, and custom integrations. It can be deployed inside of your AWS private cloud, or managed by our team. Apply here or email us at to inquire about pricing.