8 best neural networks for drawing online
Miscellaneous / / April 05, 2023
A selection of services to which access is open to everyone.
Developers continue to improve neural networks in the field of image generation. Now in the public domain there are systems that can create impressive pictures on text requests, based on pictures or schematic sketches.
The quality and detail depend on the input data - not every user will be able to immediately get a masterpiece. But these algorithms can be used, for example, to form the basis of a future illustration. Also, neural networks are suitable if you need to sort through many options for images that cannot be done only on your own in a short time.
1. midjourney
midjourney - one of the most advanced systems at the moment. The developers have so far opened access to the neural network through a server in Discord, where you need to enter the desired queries in order to get pictures. First you need to read the community rules, then go to one of the #newbies channels, enter the /imagine command in the chat and formulate your request in English.
The service generates four options: for each, you can increase the resolution or ask the neural network to redo it in the same vein. With the help of a neural network, it will be possible to create a series of images in a huge number of styles and with different levels of detail. Also, the system can combine several pictures into one.
There is a limit for free image generation - 25 pieces per user. A $10 per month subscription will allow you to make about 200 images in a dedicated channel without a long wait. An extended tariff for $30 gives access to 15 hours of neural network operation. During this time, you can get a large number of pictures.
Try Midjourney →
2. Stable Diffusion
Another popular neural network that can produce impressive results. Stable Diffusion allows you to customize the accuracy of matching the query, the number of steps, and other parameters. You can test the system through the Hugging Face service. But on this site you have to wait in long queues to make a request.
Another option is to use the DreamStudio platform, where everything works faster, but you need to create an account. You can sign in with your Google account or Discord. In DreamStudio, each user is given a limited number of credits that can be spent on generation. The more pictures and the higher the resolution, the more expensive each iteration is.
Try Stable Diffusion with Hugging Face →
Try Stable Diffusion with DreamStudio →
3. Craiyon
In Craiyon, you do not need to additionally log in and delve into the settings - after launching the site, you can immediately formulate a request in English and click on the Draw button. The result will be ready within a couple of minutes.
The neural network generates nine pictures at once according to the given description. This helps you choose the most suitable option. The service attracts users with its simple and at the same time sufficiently high detailing of images.
Try Craiyon →
4. Starryai
This neural network creates images based on a text description with additional parameters. Before starting the generation, you can choose a style, add a reference image, and also specify the number of variations and the level of elaboration.
In addition to the image generator itself, Starryai has elements of a social network. You can see other people's work here get inspired and study the most effective queries. To create images, a system of credits is used, which are not only bought, but also earned. In the second case, you will have to perform tasks like viewing ads.
The results are not always impressive - you will have to select more successful indications in the request. The system can produce both a strange abstraction and work with voluminous detailed objects. Then you can pass the finished image through the neural network again or simply increase the resolution for additional credits.
Try Starryai →
starryai
Price: Free
Download
Price: Free
5. GauGAN2
The Nvidia GauGAN2 neural network, trained on millions of nature photos, generates realistic images from a brief description. The system is also able to convert even simple outline into photorealistic landscapes.
In the test online editor, you can create pictures based on thumbnails, and then supplement the images using text queries. The tool works best with simple descriptions. He perceives complex structures with difficulty: the output is strange objects.
Try GauGAN2 →
6. dream
This is a service based on a neural network from Google that analyzes images for search queries. Dream creates abstract images for free in dozens of styles. To do this, you need to go to the website of the neural network or download the mobile application.
To start the generation, you must enter a text query up to 200 characters, as well as specify the appropriate style from the proposed ones. You can also upload an image as an example. It is also possible to add your details crypto wallet with NFT projects, on the basis of which the neural network will make new remix variations.
Try Dream →
Wombo Studios Inc.
Price: Free
Download
Price: Free
7. Kandinsky 2.0
Kandinsky 2.0 is a neural network from the developers of Sberbank. Previously, the company's engineers have already released the ruDALL-E "Malevich" and the first version of "Kandinsky", which can only be accessed in the Salyut mobile application through a voice assistant.
The neural network generates good options in two dozen styles to choose from. On the official website, you can choose the resolution of the image. For the request, you can use not only English, but also Russian, as well as combine several languages from the available 101.
Try Kandinsky 2.0 on the official website →
Try Kandinsky 2.0 on Fusion Brain →
8. This X Does Not Exist
This X Does Not Exist is a set of services that generate images and other objects in different directions and areas. Here you can get non-existent people and cats, sneakers and a starry sky.
If you do not have a specific request or time, you can go to one of the sites with a general concept. Here you don’t have to configure anything at all, enter and even click - after each page refresh, a unique image generated by the neural network is created.
Try This X Does Not Exist →
Read also💻📱
- AI Picasso will generate images using the Stable Diffusion neural network right on your smartphone
- 200 avatars by photo: the Lensa application with a neural network is storming the Internet
- A new Chinese service with a neural network turns any photo into an anime
- The Nolibox Creator web service with a neural network will create a picture from a text description
- How to reduce photo size online and offline: 8 tools