The GenAI Blog

How to Create and Customize an Image with ChatGPT and DALL-E

In this guide, we’ll walk through the process of creating a stunning image using OpenAI’s DALL-E 3 through ChatGPT-4o and then adding custom text to it, enabling you to produce personalized visuals ready for download.

---

Step 1: Generating the Image with DALL-E

1. Start with a Clear Concept: Begin by describing the image you want in detail. For instance, if you want a vibrant landscape, include elements such as colors, scenery, lighting, and any additional objects (e.g., birds, wooden pier). Example prompt: “A breathtaking and colorful original landscape featuring a serene mountain lake surrounded by vibrant autumn foliage, snow-capped peaks in the background, and a dramatic sunrise casting golden light across the scene. Include reflections of the scenery on the calm water, with a wooden pier extending into the lake for added depth. Birds in flight and a misty atmosphere enhance the dreamy quality. The style is vivid and photorealistic, designed to captivate attention and evoke awe.”
2. Use ChatGPT’s DALL-E Integration: Input your detailed prompt into ChatGPT, specifying that you’d like an image created using DALL-E. The tool generates the image according to your description.
3. View and Save the Generated Image: Once the image is ready, it’s displayed in the chat. You can download it directly in webp format by clicking on the image or using the provided link if specified.

---

Step 2: Adding Text to the Image

After generating the image, you might want to personalize it by overlaying text, like a title or a motivational quote. Here’s how you can do that:

1. Describe the Text Addition: Ask ChatGPT to add text to the image. Specify:
   • The text you want (e.g., “Magic Nature”).
   • The desired font size and color (e.g., large white text with a shadow).
   • The positioning of the text (e.g., centered, at the bottom).
   Example prompt: “Add the text ‘Magic Nature’ to the bottom center of the image in large white font with a subtle shadow for contrast.”
2. Execution Using Image Processing Libraries: ChatGPT uses libraries like Python’s Pillow (PIL) to edit the image. The process involves:
   • Loading the original image.
   • Setting the font type, size, and color.
   • Calculating the position of the text to ensure it’s properly aligned.
   • Adding optional effects like shadows for better visibility.
   • Saving the updated image.
3. Download the Edited Image: Once the image is updated, ChatGPT provides a downloadable link to the modified file in png format.

---

Practical Application of This Workflow

For this guide, here’s what we did:

1. Generated a Vibrant Landscape: Using a detailed prompt, we created an original landscape image featuring colorful autumn foliage, a serene mountain lake, and a dramatic sunrise.
2. Customized with Text: We added the text “Magic Nature” to the image in a bold, centered style with a shadow for added depth and clarity.
3. Downloaded the Final Image: After the modifications, the final image was provided with a download link for easy access.

---

Tips for Success

• Be Descriptive: The more detailed your prompt, the closer the generated image will align with your vision.
• Choose Readable Fonts: When adding text, ensure it contrasts well with the image for readability.
• Iterate: If the result isn’t perfect, refine your prompts and try again.

By combining DALL-E’s generative power with simple image editing techniques, you can create unique and personalized visuals for presentations, social media, or creative projects.