With the release of neural networks capable of generating high-quality images, the phrase: “Just a little more, and neural networks will finally replace artists,” no longer seems like a joke. Drawings created by artificial intelligence have gone from strange abstractions to impressive masterpieces. Today, neural networks can create pictures in a matter of minutes on a text request, as well as using a sketch or photograph as a reference.
The following is a list of best free neural networks onlinedrawing from photos and words. With their help, anyone can create a unique image in the style of interest based on the proposed or personally uploaded pictures.
How neural networks draw by words
With the development of artificial intelligence, neural networks are getting better at generating realistic images and other visualizations. This is achieved through deep machine learning, which involves a giant dataset collected on the basis of generative and discriminatory models:
- discriminative modeling — supervised machine learning, used for classification or regression, which distinguishes decision boundaries by inferring knowledge from existing/loaded data;
- generative modeling is an algorithm trained to generate images, текст, sound, video and others based on patterns observed in existing/loaded data.
When the neural network generates images on demand or already initially loaded art, a task is launched based on generative modeling. This process consists of the following steps:
- Preparatory. An attempt by a neural network to understand the distribution of data points.
- Starting the main process. Collection of data for their subsequent structuring.
- Providing a result. Issuance of the finished image created on the basis of the collected material.
A simple rule follows from this: if you want to get a result that will meet your personal expectations as much as possible, formulate clear requirements and / or present the most elaborate uploaded data (if text, then expanded; if image, then high-quality).
Neural networks that draw according to words and queries in 2024
For convenience, we have compiled a list of neural networks that draw according to a description or a preloaded image, fixed their main advantages and disadvantages in tables, and also attached detailed instructions for using them.
QQ World
- Site: official site.
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: Chinese.
- Query Language: this neural network does not work with requests.
QQ World has become one of the most popular sites in a short period of time due to its intuitive interface and accessible functionality. The neural network draws anime from a photo, is completely free, and also does not require additional registrations, offering free use of its capabilities.
Advantages | Disadvantages |
Ongoing support: AI developers improve their creation on a regular basis. | Daily load at the office. site due to constant influx of users. |
Uniqueness and lack of competition: is the only artist with artificial intelligence who makes anime out of photos. | Unstable PC version: if you don't want to experience slow downloads or UI lags, please visit the official website from your mobile device. |
Free to use and no mandatory registration requirements. |
How to use
Playground
- Site: official site.
- Availability: shareware program.
- Subscription costA: from $10 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Playground is a neural network that allows you to create a unique image that meets any requests and wishes. There are a huge number of presets to choose from, as well as the ability to upload personal photos and sketches, which can then be really transformed beyond recognition using various sliders, functions, etc.
“Artificial intelligence draws from photos and words,” the description of the site sounds like its main advantage, but it is not without other “pluses”.
Advantages | Disadvantages |
Availability of free functionality: limited to the ability to create up to 1000 images per day. | Mandatory registration through Google. |
A neural network that is capable of creating high-quality exclusive images both on demand and using a pre-loaded image. | Limited free functionality: can be expanded by purchasing a paid subscription (monthly update). |
Diverse functionality: there are additional sliders and settings that can captivate for long hours. |
How to use
Stable Diffusion Online
- Site: official site.
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Stable Diffusion Online has gained immense popularity along with the programs already reviewed. The neural network draws from a photo for free, visualizes the craziest text queries, and also turns children's scribbles into beautiful illustrations. In addition, it provides a program that can be downloaded and installed on a PC.
Advantages | Disadvantages |
Ongoing support: AI developers improve their creation on a regular basis. | Demanding neural network: if you want to use its functionality for some working moments, you will have to get a powerful video card with a large amount of video memory. |
The ability to install the program on a computer and use its basic functions for free. | Difficult to learn interface and lack of Russian language. |
No mandatory registration requirements. |
How to use
artbreeder
- Site: official site.
- Availability: shareware program.
- Subscription costA: from $8,99 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Artbreeder allows you to create bright and unusual portraits thanks to its wide functionality. To do this, just upload any picture, and then start editing it using the sliders and settings: age or make younger, add a mustache or apply the effect of "cartoon eyes" - the choice is huge.
Advantages | Disadvantages |
The neural network draws pictures upon request and the uploaded image, subsequently providing the ability to modify the result using sliders and various settings. | Limited free functionality: can be expanded by purchasing a paid subscription (monthly update). |
Clear and simple interface. | Mandatory registration through Google. |
An open user gallery that you can use to create your own art. |
How to use
Lens
- Site: page on Google Play, page on Apple Store.
- Availability: shareware program.
- Subscription cost: from 199 rubles per month.
- Platforms: mobile devices.
- Interface language: English.
- Query Language: English.
Lensa is a mobile application that became popular after the introduction of the "Magic Avatars" option. With its help, users can create unique avatars generated by a neural network from a photo. Internal filters allow you to turn the uploaded selfie into a "work of art" from a Disney cartoon or, for example, your favorite Japanese anime.
Advantages | Disadvantages |
By uploading only one photo, you can get from 50 unique images that "repeat" the real appearance of the user. | You can pay for a subscription only by workarounds or through intermediaries: official methods are blocked for Russian users. |
Continuous maintenance: The developers of the application improve their product on a regular basis. | Short free period, as well as mandatory registration through Google or Apple. |
Clear and simple interface. |
How to use
Dream
- Site: official site, page on Google Play, page on Apple Store.
- Availability: shareware program.
- Subscription costA: from $9,99 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Dream is a standalone application for PC and phones. Mainly used for creating vertical covers. The main styles are realism and abstractionism, which provides an opportunity to experiment and get unusual art.
Advantages | Disadvantages |
Despite the main artistic trends, the neural network creates images in different styles: during generation, you can choose the desired one yourself. | Despite the fact that most of the functionality on PC is free, on Android and iOS, much can be unlocked only after purchasing a subscription. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Mandatory registration through Google or Apple. |
Clear and simple interface. |
How to use
starryai
- Site: official site, page Google Play, page on Apple Store.
- Availability: shareware program.
- Subscription costA: from $8,99 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
The neural network draws for free from photos and words - this is the phrase that can describe the site in question, the main advantage of which lies in the chosen style directions: Argon, which is created for artistic visualization (more realistic compositions), and Altairaimed at creating fantasy and abstract art. With the help of this AI, you can generate bright images inspired by fairy tale motifs or turn a photo into an anime in a couple of clicks.
Advantages | Disadvantages |
The neural network creates images in different styles: when generating, you can choose the desired orientation yourself. | Mandatory registration through Google or Apple. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Limited free functionality: can be expanded by purchasing a paid subscription (monthly update). |
Clear and simple interface. |
How to use
Deep Dream Generator
- Site: official site.
- Availability: shareware program.
- Subscription costA: from $19 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
The first neural network that started creating art is what they say about the Deep Dream Generator in art circles. Hugely popular back in 2015, its popularity now pales in comparison to other free AI that draws from photos and requests. However, he continues to work and surprise regular users with unusual images, vaguely reminiscent of the work of Vincent van Gogh, Jimi Hendrix or, for example, Edvard Munch.
Advantages | Disadvantages |
Uniqueness and lack of competition: offers unusual stylistic directions that other popular neural networks do not have. | Mandatory registration via Google or Facebook, use of e-mail is possible. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Short free period compared to other popular neural networks. |
Clear and simple interface. |
How to use
Kandinsky
- Site: official site, Rudalle (the first version of the official website of the neural network), the Salyut voice assistant from Sberbank (through the "Turn on the artist" function in the system).
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: Russian.
- Query Language: Russian.
Kandinsky is a neural network in Russian from Sberbank. It generates images not only on demand, it can also turn into an AI that creates pictures from a photo. The main feature of the service in question is that it works great with different styles. The finished results, in turn, can be adjusted by additional clarifications on the part of the user.
Advantages | Disadvantages |
Free to use and no mandatory registration requirements. | The lack of a single platform where all the functionality would be collected. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Medium level of sophistication of the program: at times, algorithms misunderstand and implement complex or unusual requests. |
Variety of styles: implemented more than 20 artistic directions. | The quality of the generated art is inferior to some popular analogues (for example, Midjourney). |
Possibility to enter a query up to 2000 characters long. |
In a separate guide, you can find work at Kandinsky in more detail: we figured out how to use the neural network, correctly compose prompts, upload your photos, etc.
How to use
ruDALL-E
- Site: official site.
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: Russian.
- Query Language: Russian.
ruDALL-E is a Russian free neural network that draws by words. Kandinsky's subsidiary service, which uses the back diffusion method and, after saving the text query, generates colorful images: up to 9 pieces with the ability to resize. There are many filters and directions: the creators of the site have tried to ensure that each user finds what he likes.
Advantages | Disadvantages |
Generates up to 9 images and then saves the edit. features (change the size, color scheme, etc.). | The lack of a function for downloading a reference image: it will not be possible to call ruDALL-E a “free neural network for photos” soon. |
Free to use and no mandatory registration requirements. | Blurring of some final art. |
Clear and simple interface. | Lack of diversity in the preparation of "narrow" queries. |
midjourney
- Site: official site, official server in Discord.
- Availability: since March 30, 2023 developers canceled free access.
- Subscription cost: from 1900 rubles.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
To date, Midjourney is considered one of the most progressive free neural networks that draw from photos and words. Despite the complex mechanics and the abundance of all kinds of parameters, commands and settings, working with the AI in question involves obtaining high-quality art for any prompt. The base of 10 million contributors who consistently generate amazing images for personal use is proof of that.
Advantages | Disadvantages |
Functionality that allows you to generate unique and high-quality art. | You need a Discord account to use. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Requests are accepted exclusively in English: prompts in other languages are acceptable, but may lead to a result that is completely inconsistent with the desired one. |
Creates images in different styles: when generating the desired orientation, you can choose yourself by writing the appropriate request in the "promt" box. | Difficult to use interface. |
The neural network draws according to words, as well as according to pre-loaded pictures (a link to a photo or art is fixed in the “promt”). |
In a separate guide, you can find work at Midjourney in more detail: we figured out how to use the neural network, correctly compose prompts, upload your photos, etc.
How to use
Masterpiece
- Site: official site, page on Google Play, page on App Store.
- Availability: for free.
- Subscription cost: not.
- Platforms: mobile devices.
- Interface language: English and Russian.
- Query Language: English and Russian.
Masterpiece is an application created by the Russian company Yandex, which is available for download on Google Play and the App Store (also supported by emulators). The main feature of the neural network is the creation of images based on a preloaded image or a point query. Subsequently, the arts lend themselves to detailing thanks to the corresponding text sentences entered both in English and in Russian.
Advantages | Disadvantages |
Free use with the ability to formulate requests in Russian. | Mandatory registration through Google or Apple. |
Clear and simple interface. | Poor quality image generation with people and animals. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Low level of sophistication of the program: at times, algorithms misunderstand and implement complex or unusual requests. Some words may also be "unfamiliar" to the AI. |
An open user gallery that you can use to create your own art. |
In a separate guide, you can find work in Masterroom in more detail: we figured out how to use the neural network, correctly compose prompts, upload your photos, etc.
How to use
kaiber
- Site: official site.
- Availability: shareware program.
- Subscription costA: from $10 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Kaiber is a unique neural network capable of creating animated videos from images: even the Linkin Park group turned their attention to this program, preparing a video for the song “Lost” with its help. The AI in question can greatly simplify the preparation of an exclusive video in anime style: many artists learn to animate their creations for months, and Kaiber allows you to master such a difficult skill in a couple of clicks.
Advantages | Disadvantages |
Artificial intelligence draws online animated videos of different time durations according to pre-loaded images and a detailed text query, which makes it a unique representative among neural networks. | Users are given only 30 free seconds to create an anime video. Having decided to use AI, think over the concept of the video in advance if you don’t want to overpay for extra. time. |
Ongoing Support: The developers of the app improve their product on a regular basis. | Difficult to learn interface: due to the limited amount of time provided for free, it is recommended to learn how to use the neural network in advance so as not to spoil the result. |
A sudden change of frames can lead to blurring. |
How to use
crayon
- Site: official site.
- Availability: shareware program.
- Subscription costA: from $5 per month.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
This neural network draws pictures on request, providing 9 high-quality images as a result, where each one is different from the previous one. The styles are also conceptually different: the reason is that the developers are trying to give the user a different view of fixed prompts in order to subsequently reach an exclusive result. Uniqueness is the main credo of the creators.
Advantages | Disadvantages |
Conditionally free to use and no mandatory registration requirements. | Medium level of sophistication of the program: at times, algorithms misunderstand and implement complex or unusual requests. |
Ongoing Support: The developers of the app improve their product on a regular basis. | There is no way to generate images from photos. |
Clear and simple interface. |
How to use
DALL-E Mini
- Site: official site.
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
DALL-E Mini is a sister service of Craiyon and a typical representative of its niche, which provides access to creating personal visual notes to invented text phrases and expressions. It is considered one of the most popular AI as it is easy to learn and has an intuitive interface.
Advantages | Disadvantages |
The presence of an acceptable number of free attempts to generate personal requests and the absence of mandatory registration requirements. | After the end of the free "period", any attempts to continue using the neural network lead to a notification from the official website about the transition to Craiyon. |
Ongoing Support: The developers of the app are constantly improving their product as they want to attract more users to Craiyon. | Medium level of sophistication of the program: at times, algorithms misunderstand and implement complex or unusual requests. |
Clear and simple interface. |
Scribble Diffusion
- Site: official site.
- Availability: for free.
- Subscription cost: not.
- Platforms: computers and mobile devices.
- Interface language: English.
- Query Language: English.
Scribble Diffusion is an easy-to-use neural network that will be a real lifesaver for ordinary users who like to format sketches into high-quality images (be it a photo or simple art). To do this, just draw the desired object in the square window, fix the request in English in a special field, and finally click "Go".
Advantages | Disadvantages |
Free to use and no mandatory registration requirements. | Blurring of the final art for complexly formulated requests. |
General accessibility: to please those who just want to relax and have fun. | Limited functionality: only two actions are available to the user - draw an object and compose a prompt. |
Clear and simple interface. |
Don't expect anything serious from Scribble Diffusion, as it's made for fun, not for permanent professional use.
Read on:We hope that with the help of a selection from the article you were able to find a suitable free neural network that draws from photos and words. Write to commentsIf you have any questions, the author will try to help.
All these neural networks are crap and they are of no use. The images are crooked or not at all what you need!
TOP
I don't know why it doesn't work
Class
Understand the human word
Thank you, great breakdown. Especially for details, links.
I sit and try. I want to learn how to transform a person from a photo into a different image. I can’t master Mijorni yet, it’s something dense: go there, I don’t know where, bring something, I don’t know what. Kryon doesn’t understand me at all, if you write “knight” it looks like bastards in caps... But I will continue to try to get along with others. Time is also a pity - hanging around there is no use.
Your site is interesting, thank you.
Thank you for such an extended comment! We are very happy with the feedback from the audience. With:
As for Midjourney: we have a detailed guide to using this neural network. We recommend that you read, dismantled many points. And so we wish you good luck in mastering neural networks!
Thank you
from your list of FREE, mostly all Paid, and some do not work at all, or the host is not found.
Hello! The author of the article tested all the neural networks presented in the guide. Each of them works correctly, there are no errors in the system of this or that service.
Links that lead to the official website of a particular neural network are also correct. As for free/paid content: some neural networks have paid features inside, but at the same time they are not devoid of free content that can be tried by ordinary users.
The Sber network is rather stupid in places (compared to the same "Craiyon"), and there is a big problem with the variability of the results. Yes, and a constantly crooked pop-up frame for selecting image zones when trying to save - well, such a thing.
The article is excellent, all the pros and cons are painted.
Very interesting and informative.
cat robot
in which neuron you can enter your photo and what would it generate a picture
in migiorni)
Dream by WOMBO perfectly understands prompts in Russian too, you just don't need to write essays in 200+ characters a la Leo Tolstoy. In general, in my purely subjective opinion, Midjourney is in the first place, Dream by WOMBO is in the second, and Deep Dream Generator is in the third. The rest is still growing and growing until the normal generation of pictures.
me to take a step
Everything is in English.
Everything is paid.
25 minutes free is ridiculous. Animation should be solid, not just cartoons.
Few pictures to choose from.
All benefits only to corporations, and users only pictures to them,
photos are carried in the goiter, moreover, for a fee.
How can I start drawing?
What should be written in the website?
krasava
Thank you very much for your work. !
Thank you for your kind feedback, we're glad to try!
Wow.
Super class!
Thank you. I am now listening to a webinar - a presentation of courses on training the profession of "neuroillustrator". Sells similar information for good money.)))
We are happy to help 🙂 By Midjourney, we tried to analyze the main points in detail, but we will still supplement with tips, we will share our personal experience a little later
IMHO, but the highest quality now is Midjourney.
In the near future we will write a separate article on it and add a link here. It is of better quality, but there are a little more troubles with the use, so we will analyze everything in detail in a separate instruction.