Detailed guide on creating AI images using Midjourney and ChatGPT

“This tutorial will help you create unique and artistic AI images using Midjourney and ChatGPT. Explore and experience the power of artificial intelligence (AI) in the era of Industry 4.0.”

Introduction

Midjourney is one of the most popular AI image creation tools today. It uses a GAN (Generative Adversarial Network) model to generate new and unique images from input data. ChatGPT is an automatic language model with the ability to understand and generate natural language responses. The combination of Midjourney and ChatGPT allows us to create new images with high realism and creativity. Let’s take a look at the detailed guide on how to create AI images with Midjourney and ChatGPT in this article.

What is Midjourney?

Midjourney is an independent AI art image creation tool that can turn text prompts into images, similar to DALL-E OpenAI and Stable Diffusion. All you need to do is enter a few words, and this tool will provide you with an AI-generated original image using machine learning. Midjourney is different from other AI art creation tools as it can only be accessed through a Discord bot. To create an image, you must message the bot directly on the official Midjourney Discord page or by inviting the bot to a third-party server. Mid Journey is open for community use. Anyone with a Discord account can join the Midjourney server and start using it for free. When you are satisfied with what the bot returns, you can upgrade the image and download or send it to yourself.

Versions of Midjourney

As of March 2023, Midjourney has released 5 official versions along with various trial versions. The latest version, V5, with the most advanced features, was launched on March 15, 2023.

Midjourney V.1: released on February 22, 2022, and used as the default model until April 22, 2022. Version 1 produced very abstract images and was suitable for artists, but had relatively low resolution.

Midjourney V.2: launched on April 22, 2022, and used until July 22, 2022. MJ V.2 was more creative and produced unique, colorful images with patterns, but still had low image quality.

Midjourney V.3: continued to replace version V.2 and was used as the default for MJ until November 22, 2022. With this version, AI creations with high creativity were possible, and the color layout was appropriate.

Midjourney V.4: had a completely new code base and AI architecture designed and trained on the new Midjourney AI cluster. Version 4 had more knowledge of living things, places, objects, etc. This version had three “styles,” 4a, 4b, and 4c, which were slightly different with minor adjustments in the model’s style. V.4 was able to produce more vivid and realistic color and detail images than previous versions, but natural language processing and creative layout were not yet perfected.

Midjourney V.5: this is the latest and most advanced version to date, with high consistency and superior performance in interpreting natural language prompts, higher resolution, and advanced features such as repeating to create seamless patterns for fabrics, wallpapers, and patterns.

In addition, MJ has other models such as Niji, a collaboration between Midjourney and Spellbrush, which was adjusted to create anime and illustration styles. MJ test model was used for testing purposes.

Currently, Midjourney is open for free use in the community but with a limit on the number of image creations per Discord account. To create more images, you need to upgrade to their paid packages. Details can be found in the later part of this article.

What is the Discord application?

Discord is a popular voice and text chat application used by gamers and online gaming communities. It allows users to create servers to chat, share documents, images and videos, or chat through voice or video with others.
This application is designed to be user-friendly and support multiple users chatting online in real time. It is widely used for activities such as gaming, online meetings, remote teaching, and exchanging information in interest groups.
Discord also provides security features such as encryption, access restrictions, and account management to ensure the safety of users.

Why do we need to learn about Discord?

It’s because Midjourney uses a chatbot on Discord to interact and create AI images. You have to call out commands and input prompt messages so that the bot can return AI images as requested.

Why should we combine Midjourney and ChatGPT to create AI images?

Midjourney is an AI processing tool used to create images based on text descriptions. To produce accurate, beautiful, and desired outputs, you need to have good input. This is where the power of ChatGPT comes in. ChatGPT can provide natural language processing and understand synonymous meanings, helping to create the most complete and detailed prompt sentences. From there, the quality of the AI images generated will be accurate and intuitive for users.

Detailed guide to creating AI images with Midjourney

Join Midjourney’s chat server on Discord

Create a Discord account

To use Midjourney, you first need to register for a free Discord account. The Discord chat application can now be used on all platforms such as Windows, Web browser, Android, IOS, … Learn more about how to create a Discord account here.

Using commands to create the first image

You need to interact with the Midjourney Bot on Discord using Commands. Commands are used to create images, change default settings, track user information, and perform other useful tasks. The /imagine command creates a single image from a short text description called a prompt.

Now let’s try creating your first image with the following steps:

  1. In the chat box, type /imagine prompt: or select the /imagine command from the popup window.
  2. Enter the image description you want to create in the prompt field. For example: a cat
  3. Press enter to send your request and wait for the bot to respond. Midjourney Bot will return a result consisting of 4 different variant images.

Some notes when creating images:

Customizing and Upgrading Images

Once the four generated images are ready, you can choose to upgrade the images or create additional variations of one of the options using the rows of buttons below.

To upgrade an image:

  1. Look for the first row of buttons below the image grid with the characters U1-U2-U3-U4. These U buttons will upgrade the selected image to a larger version with more details.
  2. Press the U buttons corresponding to the images you want to select, with the numbers corresponding to the images from left to right, top to bottom.
  3. The image upgrade process is almost complete. However, if you are not satisfied with the result, a new set of buttons below will provide you with other options, such as:
    Make Variations: create a variation of the upgraded image and create a new grid of four options.
    Beta/Light Upscale Redo: redo the upgrade using a different upscale model.
    Web: open the image in your collection on Midjourney.com.
    Emoji: click the smiley face button to rate any upgraded image on the Midjourney website or in Discord. Every day, the top 1000 rated images will receive a free hour of MJ GPU.

Different variations options in different Midjourney versions:

  1. Use the next row of buttons V1-V2-V3-V4: These V buttons will create small variations of the image with overall layout and style according to the version you choose.
  2. Click the V buttons for the version you want.
  3. The bot will provide you with four new images. Follow the steps above to upgrade the image you like. Save or share your AI images Once you have created an AI image that you are satisfied with, the next step is to save it and share it with everyone. You can do the following:
    Save the image: Click on the image to open it at full size, then right-click and select Save image. On mobile phones, press and hold the image and then press the download icon in the top right corner.
    Send to yourself: You can also take a copy of the image by sending it to yourself on Discord. In the bot’s response message, click on the emoji icon, then select the envelope icon. You will find the image in your Discord inbox.

Some common commands and parameters used in Midjourney

Common Midjourney commands:

Common Midjourney parameters (parameters are always added at the end of the command prompt. You can add multiple parameters to each prompt.):

Reference materials for using Midjourney

To use Midjourney for creating AI images for beginners, you should refer to the information sources and libraries on the internet to avoid wasting GPU and save time. Below are some websites for you to learn more:

Conclusion

Drawing or creating images with AI is a trend in the current 4.0 era. Hopefully, the article “Detailed guide to creating AI images with Midjourney and ChatGPT” from AIoutils.com will help you create beautiful and creative images. In addition, you can also learn about many other AI image creation tools such as Stable Diffusion or DELL-2 from OpenAI to experience and compare the differences between them. Wish you success in creating your own AI images.

Exit mobile version