MovieMinter v2 - Part 1: Setting up Replicate

In this article series, I will be using LivePeer, Replicate and Replit to create a React-based Decentralized Generative AI Movie NFT Minter. In this article, I will walk through the process of creating a Replicate deployment to act as the AI-backend.

In the first article, we will learn how to set up a Replicate deployment to use a popular text-to-video model.

What is Replicate?

Replicate is a service that offers easy access to many popular AI models as they are released. They host a wide range of models useful for different purposes including generating images, voice, text-to-video and more.

The backend AI model I am going to use for my MovieMinter v2 project is damo-text-to-video.

damo-text-to-video allows you to add a prompt, configure some details and then create a custom-generated video on demand using generative AI.

Getting Started With Replicate

If you haven’t yet created a Replicate account, please do so.

Next, navigate to this area and create an API token. Name your token whatever you want.

Next, you will need to create a “Deployment” in order to make API calls using your token to the damo-text-to-video endpoint. There will be some instructions around details of your deployment, fill those in as desired.

Once the deployment is created, you will now be able to make API calls to Replicate from your back-end. The deployment will appear under Dashboard -> Deployments as below.

Under the deployment, there will be details around how to make API calls to this deployment on demand with some boilerplate.

Once you have your deployment, now we are ready to build the backend of the application.