How to use Dall-E 3 with Chain of Thought Prompting

I’m super excited to delve into a topic that’s a tantalizing blend of creativity and technology: using Dall-E 3 with a “Chain of Thought” prompting method.

I felt compelled to share my own experience and walk you through the complex, yet fascinating, principles behind it. So, let’s dive in, shall we?

Read more or watch the YouTube video(Recommended)


What Is Dall-E 3?

The Basics

Before we get to the juicy part, let’s quickly cover what Dall-E 3 is. It’s an extension of OpenAI’s text-to-image model, Dall-E, designed to create highly detailed and contextually accurate images from textual prompts. What sets Dall-E 3 apart is its refined architecture and enhanced capabilities, making it more versatile and accurate than its predecessors.

UPDATE: We now have access to Dall-E 3 via the API, check it our in action here.

The Power of Prompt Engineering

One of the most compelling aspects of Dall-E 3 is the freedom it provides with Prompt Engineering. Essentially, you can instruct Dall-E 3 to generate images by feeding it a series of text prompts. The quality and creativity of the output often hinge on how well you craft these prompts. And this brings us to the central theme of today’s discussion: Chain of Thought Prompting.

picture of dall-e 3 chain of though prompting

The Principle of Chain of Thought Prompting

Lets take a look at how I use the Chain of Thought principles to create a better experience of creating images with Dall-E 3 highly influated by the set custom instructions system prompt.

You can find the system prompt and a prompt example of this method here.

Crafting the Perfect Prompt

The idea behind Chain of Thought Prompting is to decompose your requirements into a series of smaller, manageable ideas created by the set ChatGPT System Prompt. For instance, if you’re designing a YouTube thumbnail, you might break down your prompt into style, text, and objects. This detailed list guides Dall-E 3 in a step-by-step fashion to craft an image that ticks all the boxes.

Structuring the Process

In the Chain of Thought approach, it’s advisable to start by creating a long, detailed list of individual ideas. These could range from deciding the format (say, 16:9 for YouTube thumbnails) to the color palette and aesthetic style. The process helps in crafting a detailed and layered prompt that guides Dall-E 3 to create a multi-dimensional image.

Example in Action

For instance, consider designing a thumbnail with a 90s retro hacker style. The big text would say, “You’ve been hacked,” and the image would include a vintage computer running green code. By breaking it down into style, text, and objects, you craft a layered prompt that Dall-E 3 can navigate to produce a thumbnail that isn’t just eye-catching but also rich in context and detail.

picture of dall-e 3 chain of though prompting

Why Chain of Thought Prompting Boosts CTR

So, why does this method work so well, especially for purposes like creating YouTube thumbnails aimed at high click-through rates (CTR)? The answer lies in the richness of detail. By elaborating on each aspect, from style to objects and text, you can tap into the viewer’s nostalgia or create a compelling narrative that makes the thumbnail irresistibly click-worthy.

My Personal Experience

I put this methodology to the test by designing a series of thumbnails and even personal cards. The level of detail and the uniqueness of each design were staggering. Whether it was the moodier atmosphere of a 90s arcade game or the vibrant colors of an anime-themed thumbnail, Dall-E 3 delivered beyond my expectations. And all it took was a well-crafted, detailed prompt following the Chain of Thought approach.

picture of dall-e 3 chain of though prompting


In the ever-evolving landscape of AI and machine learning, Dall-E 3 stands as a testament to how far we’ve come. The Chain of Thought Prompting technique serves as a fine example of how the fusion of human creativity and machine intelligence can produce awe-inspiring results. It’s not just about instructing a machine to perform a task; it’s about collaborating with it to bring an abstract concept to life.

So, if you’re intrigued by the boundless possibilities that Dall-E 3 offers, I highly recommend giving Chain of Thought Prompting a try. You’ll be amazed at how a simple yet detailed text prompt can translate into a visually stunning and emotionally resonant image.


What is Dall-E 3

Dall-E 3 is an advanced text-to-image model by OpenAI that takes textual prompts to generate highly detailed and contextually accurate images. It’s an extension of the original Dall-E but comes with a refined architecture and enhanced capabilities, making it more versatile than its predecessors.

What is Chain of Thought Prompting in Dall-E 3?

Chain of Thought Prompting is a methodology used to craft detailed and layered text prompts for Dall-E 3. By breaking down your requirements into smaller, manageable ideas, you can guide the model to produce images that are not just visually appealing but also rich in context and detail.

How can I use Dall-E?

If you are a ChatGPT plus user, you can use Dall-E in the dropdown meny on the ChatGPT webbrowser.

Leave a Reply

Your email address will not be published. Required fields are marked *