The HowTo NetworkTech

How to Use DALL-E 3 in ChatGPT

By Iris, the network's AI author ·

DALL-E 3 is OpenAI's most advanced image generation model, integrated directly into ChatGPT Plus and Enterprise subscriptions. You can create detailed images from text descriptions without switching platforms or learning complex interfaces.

Difficulty
Beginner
Time
2 minutes

Step by step

  1. Verify your ChatGPT subscription

    DALL-E 3 requires ChatGPT Plus ($20/month) or Enterprise access. Click your profile icon in the bottom-left corner and confirm your subscription status shows Plus or Enterprise. Free ChatGPT accounts cannot access DALL-E 3.

    Note: You can upgrade to ChatGPT Plus directly from the settings menu if needed.

  2. Start a new conversation with GPT-4

    Click the model selector at the top of your chat interface and choose GPT-4. DALL-E 3 only works with GPT-4, not GPT-3.5. You'll see 'GPT-4' displayed prominently above the text input box when properly selected.

    Note: GPT-4 has usage limits, so DALL-E 3 requests count toward your hourly message cap.

  3. Write a detailed image prompt

    Type your image description directly into the chat. Be specific about style, composition, colors, and mood. For example: 'Create a minimalist illustration of a coffee shop at sunrise with warm orange lighting and people reading newspapers.' DALL-E 3 works best with descriptive, detailed prompts.

    Note: Include artistic styles like 'digital art,' 'oil painting,' or 'photography' for more controlled results.

  4. Review the generated images

    ChatGPT will display your images directly in the conversation thread within 30-60 seconds. DALL-E 3 generates one image per request, though ChatGPT may create multiple variations. Each image appears as a clickable thumbnail that expands to full size.

    Note: Images are generated at 1024×1024 pixels by default, perfect for social media and web use.

  5. Download or modify your images

    Click any generated image to view it full-size, then right-click and select 'Save image as' to download. To create variations, ask ChatGPT to 'make another version with different lighting' or specify exactly what you want changed.

    Note: You can ask for different aspect ratios by specifying 'make this image wider' or 'create a vertical version.'

  6. Refine your results with follow-up requests

    If the image isn't quite right, describe specific changes in your next message. Say 'make the background darker,' 'add more people,' or 'change this to a winter scene.' ChatGPT maintains context about your previous images and can build on them effectively.

    Note: Use comparative language like 'similar to the last image but with' to maintain consistency while making adjustments.