There is no doubt that DALL-E 2 and Stable Diffusion are the most popular AIs when it comes to create images from text. However, there is one that is just as powerful as the ones already mentioned and is even easier to use. This is Midjourney, the AI with which the baby dinosaurs meme was created and whose latest version has just been announced: Midjourney V5. Would you like to try it? Well, below we explain how to do it.
These are the news of version 5 of Midjourney, the AI to generate images from text in Discord
Midjourney is an artificial intelligence that allows you to generate images just by typing a textual description of it. In addition, it is one of the few AI that offers free access to any user through Discord. All you have to do is join their beta tester program and you’re good to go.
Well, the Midjourney team has just announced on their subreddit the arrival of the new version of this artificial intelligence: V5. AND… What’s new? Well, according to its developers, now the AI is more precise when interpreting the user’s instructions. Below, we summarize all the improvements presented:
- improved performance in understanding the user’s instructions.
- Greater precision when creating teeth (quantity and rows) hands (orientation and number of fingers) and faces (textures, brightness of the eyes and naturalness of the hair).
- new higher aspect ratios at 2:1 (experimental function).
- images to double resolution offered by V4 (1024 x 1024 pixels) and improved HDR.
- new parameter “–iw” for prioritize image prompts over text prompts and thus avoid unwanted text in the created image.
- new parameter “–tile” to create images that can be used as mosaics or to create patterns (experimental function).
It is necessary to mention that, for now, Midjourney V5 is only available to trial members. And unfortunately, due to the great popularity of the AI, the places to access the test of this new version have already been exhausted.
However, it is very likely that Midjourney V5 will be available to everyone for free in the next few weeks, as is V1, V2, V3 and V4. For now you can enjoy the older versions of the AI… Would you be interested in trying Midjourney? Well, below we explain how to do it.
How to use Midjourney V5 for free
How to use Midjourney V5 for free
If you are already a user of this artificial intelligencePlease know that using Midjourney V5 for free is just as easy as using the previous AI version (Midjourney V4). In fact, the only thing that changes is that now you will have to add the parameter «–v 5» so that with your textual indication the AI uses its fifth version instead of V4 which, for the moment, continues to be the default version of Midjourney.
However, if it’s your first time with Midjourney, you don’t have to worry, using this AI is very easy. The first thing you have to know is that it is necessary to have a Discord account and the app of this messaging platform on your mobile or PC. You already have it? Well then follow these steps:
- Enter the website de Midjourney (with this link) on your mobile, computer or device where you have your Discord account open. Once you are on the web page, you have to click on Join the Beta.
- The next thing will be to press Accept invitation, which will automatically open the Discord app. There this message will appear again, so you have to press again on Accept invitation.
- Doing this will add the midjourney server to your Discord, to which you will have access while it is operational (sometimes the server is offline due to the high demand for images it receives). Now you must press on the button of the three horizontal lines which is in the upper left corner.
- The next step you should do is enter any of the channels called newsbies. In this tutorial we enter the “newbies-44”. Note that any of the channels with this name are valid for using the AI.
- Once you are on the channel you will see a lot of AI generated images. To create your own image you have to write the command /imagine (prompt) in the chat and then a precise description of the image you want Midjourney to create for you.
- The textual description of the image must write it in english. The more details you provide, the more accurate the AI will be. For example, you can describe the artistic style (photorealistic, “comic”, renaissance paintingetc.), the object or subject of the image, the background, the angle, the perspective, the colors, textures, etc.
- After adding a description of the image, you can add parameters to modify technical details of the image: –ar (aspect ratio), –no objects (to remove objects from the image, example: “–no plants”), –v (AI version, example: v– 5 or v– 4), etc. Here we leave you the official list of Midjourney parameters.
- When you have written the description of the image you want, now you must send the message in the channel. When doing it for the first time, you will receive a notice about the terms and conditions of use of the AI in the chat. Here you have to click on Accept ToS to continue.
- After a few minutes, you will receive a message in the chat with 4 images that will be rendered in real time (you must wait until they reach 100%). Once the AI completes generating the image, it will appear to you 4 U-buttons and 4 V-buttons under the images. Each represents an image. The “V” buttons are used to obtain variants of the image and the buttons “OR” for download high resolution image. If you do, you’ll receive a message with the image you selected, which you can download the same way you download any other image on Discord.
And you… What are you waiting for to create your own images with Midjourney?
