テキストプロンプトから精細な画像やイラストを生成するAIアートサービス
AI画像ツール 人気のAI画像ツール midjourney.com
What is Midjourney
Midjourney is an AI image generation service from the San Francisco research lab of the same name, led by Leap Motion co-founder David Holz. It entered open beta on 12 July 2022. Type a description (a prompt) and it creates images, and it can also turn images into videos. You can use it at midjourney.com or in Discord.
What it can do
- Text to image: write a prompt in the Imagine bar on the Create page, and each run produces a set of four images
- Further editing: upscale, vary, zoom out, pan, and edit parts of an image in the Editor
- Image to video: animate Midjourney images or your own
- Personalisation: moodboards and personalisation profiles steer results towards your style
- Settings for aspect ratio, stylisation, Raw mode, Draft mode and more
Pricing
A subscription is required. The official plans (monthly / annual equivalent per month):
- Basic: US$10 / US$8, about 3.3 hours of Fast GPU time a month
- Standard: US$30 / US$24, 15 hours of Fast time and unlimited images in Relax mode
- Pro: US$60 / US$48, 30 hours, unlimited images and SD video, plus Stealth mode
- Mega: US$120 / US$96, 60 hours
Annual plans save 20%, and extra Fast time costs US$4 an hour.
Who it's for
- Illustration, concept art, posters and social media images
- Designers who care about aesthetics and a consistent style
Good to know
- Subscribers can use their images and videos in almost any way, but companies with more than US$1 million in annual revenue must buy the Pro or Mega plan.
- Only Pro and Mega include Stealth mode; on other plans your work is public by default.
- Subscriptions renew automatically, so cancel if you stop using it.
Similar tools
Compare Stable Diffusion (open source and self-hostable), Jimeng and Kling AI.