A Beginner's Guide to the Dalle-Mega Text-to-Image model

Let's learn how to run the Dalle-Mega model and turn our text prompts into images

A Beginner's Guide to the Dalle-Mega Text-to-Image model
Example output of the Dalle-mega model

Have you ever imagined a tool that could transform your written ideas into images? Sounds fascinating, right? Dalle-Mega, an AI model developed by Nicholas Celestin as part of the DALL-E playground project, does exactly this, making it an excellent companion for designers, artists, or just about anyone looking to manifest their creative thoughts visually.

Subscribe or follow me on Twitter for more content like this!

In this guide, I will introduce you to the Dalle-Mega model, ranked #190 on AIModels.fyi, that takes a text prompt as input and generates an image. I will explain how the model works, its inputs and outputs, and guide you step-by-step to use it effectively. Additionally, we'll explore how you can utilize AIModels.fyi to find similar models, helping you make informed decisions. Ready to unlock your creativity? Let's begin!

About the Dalle-Mega

Dalle-Mega, created by Nicholas Celestin, is an AI model that uses the power of AI to generate images from text prompts. The model uses an Nvidia T4 GPU to run and has an average completion time of 45 seconds per run. The cost per run is $0.02475 USD. Despite its relatively recent introduction, Dalle-Mega has already been run 11,840 times. To learn more about the model, visit the model's detail page.

Despite the availability of the Dalle-Mega model, the creator recommends using the Dalle-Mini model instead, deeming it superior. Dalle-Mini is a smaller and faster version of the model, and there is an implementation by Boris Dayma available on Replicate. It can be accessed here, and functions similarly to the Dalle-Mega model.

Understanding the Inputs and Outputs of the Dalle-Mega

Before we jump into how to use Dalle-Mega, it's essential to understand what goes in and what comes out of it.


The model takes two inputs:

  • prompt (string): This is the text or idea that you want to transform into an image.
  • num (integer): This specifies the number of images to generate. Its default value is 1.
  • model_size (string): Specifies the size of the model. The allowed value for this is MINI.


The model's output is an array of strings, where each string is a URI of the generated image.

Now that we understand what Dalle-Mega needs and returns, let's dive into how to use it.

A Step-By-Step Guide to Using Dalle-Mega

For those who prefer a hands-off approach, you can interact directly with the Dalle-Mega's demo on Replicate via its UI. This link to the demo allows you to interact directly with the interface, playing with the model's parameters, and getting quick feedback. However, if you're more of a hands-on coder, the following guide will show you how to interact with the model's Replicate API.

Step 1: Installation

Start by installing the Node.js client:

npm install replicate

Step 2: Authentication

Next, copy your API token and authenticate by setting it as an environment variable:

export REPLICATE_API_TOKEN=r8_*************************************

Step 3: Running the Model

Then, run the model:

import Replicate from "replicate";

const replicate = new Replicate({
  auth: process.env.REPLICATE_API_TOKEN,

const output = await replicate.run(
    input: {
      prompt: "..."

You can also set a webhook URL to be called when the prediction is complete:

const prediction = await replicate.predictions.create({
  version: "70dbc2347bdc08d78af32a0fca8c1cafd3a6a7c29ea2bfa5a2fbbf0e70a1e1b8",
  input: {
    prompt: "..."
  webhook: "https://example.com/your-webhook",
  webhook_events_filter: ["completed"]

This allows you to receive updates about your prediction.

Taking it Further - Finding Other Text-to-Image Models with AIModels.fyi

AIModels.fyi is an excellent resource for discovering other AI models like Dalle-Mega. The database lets you explore and compare a myriad of models, making it easy to find ones that align with your needs.

Step 1: Visit AIModels.fyi

Head over to AIModels.fyi to begin your search for similar models.

Search for models using keywords such as "Text-to-Image." This will display a list of related models.

Step 3: Filter the Results

On the left side of the search results page, you'll find filters that can help narrow down the list of models. You can sort by model type, cost, popularity, or even specific creators.


In this guide, we took a deep dive into Dalle-Mega, understanding its inputs, outputs, and using it to generate images from text prompts. We also explored how to leverage AIModels.fyi to find similar models, broadening our horizons in the world of AI-powered creativity.

I hope this guide has inspired you to harness the power of AI for your creative projects. Don't forget to subscribe to Notes.AIModels.fyi for more tutorials, updates on new and improved AI models, and a wealth of inspiration for your next project. You can also follow me on Twitter for more insights and discussions on AI. Happy creating with AIModels.fyi!

Subscribe or follow me on Twitter for more content like this!