機械翻訳 / Machine-translated
@aifriends
AI Friends（https://aifriends.jp）のクロスポスト公式アカウント。AIツールの紹介・使い方・できることを、中学生でもわかるやさしい日本語で届けます。
"I've heard of Midjourney, but what can it actually do?" If that's what you're wondering, this article is for you. Here, we'll explain the buzzworthy AI image generation tool "Midjourney" in simple terms that anyone can understand.
Midjourney is an AI image generation tool that creates high-quality images simply from text instructions. It was launched in the United States in 2022 and is now used all around the world.
Type in something like "a cat sitting by a seaside at sunset," and the AI will automatically draw that scene for you. Even without any drawing skills, you can produce professional-grade images in just a matter of seconds.
In April 2025, the latest version "V7" was released. The texture of human skin and fabric has become more realistic, reaching a level where it's almost indistinguishable from a photograph. Development of "V8," aimed for 2026, is also currently underway.
Midjourney comes packed with useful features. Here, we'll introduce five of its most notable ones.
Simply enter a text description (a prompt) to create beautiful, artwork-quality images. It can handle any genre — landscapes, portraits, abstract art, and more. You can give instructions in Japanese, but English tends to produce more accurate results.
Using "Draft Mode," added in V7, you can generate images 10 times faster than normal. On top of that, it uses only half the cost (GPU time). It's perfect when you want to experiment with lots of ideas.
This feature lets you give instructions simply by speaking into a microphone. Japanese voice input is also supported. It's handy for those who aren't comfortable with typing, or when you want to operate hands-free while on the go.
After you rate around 200 images, Midjourney learns your preferences. From then on, it automatically generates images in a style tailored to your taste. In V7, you can create multiple profiles, making it possible to switch between them depending on the purpose.
The "V1 Video Model," released in June 2025, lets you create videos up to 20 seconds long from existing images. It's still in beta (a testing phase), but it has opened the door to creating moving content, not just still images.
Even beginners can get started easily. Follow these four steps to begin generating images.
Go to the official website (midjourney.com) and create an account. All you need to do is set an email address and password. You can also use a chat app called Discord, but the web version is recommended for beginners.
Midjourney does not currently offer a free trial. You will need to sign up for a paid plan. The most affordable Basic Plan starts at $8 per month when billed annually (approximately ¥1,200). Payment is easy with a credit card.
Type in a text description of the image you want to generate. For example, writing something specific like "Mount Fuji at dusk, cherry blossom petals drifting in the wind" will yield better results. Start with short sentences at first, and as you get more comfortable, add more detail.
If you're happy with the generated image, click the download button to save it. In the web version, you can also right-click the image and select "Save As." Please note that a paid plan is required if you intend to use images commercially.
Every tool has its strengths and things to watch out for. Understand Midjourney's characteristics and make the most of it.
Midjourney is especially well-suited for the following types of people:
On the other hand, if you want to use it completely free of charge, or if you want precise control over every detail, it may be worth exploring other options as well.
In this article, we explained the AI image generation tool "Midjourney." Let's review the key points one more time.
With Midjourney, anyone can easily create professional-grade images. Why not start by creating an account on the official website and bringing your ideas to life visually?
This article is a cross-post from AI Friends.
