I've seen a lot recentlyAI DrawingRelated scenario application reports, such as AI making children's picture books, AI making personal portraits, AI making posters, AI making PPT illustrations, AI making comics, etc. The production of these scenarios all involve functions such as text-generated images, image-generated images, image expansion, and image modification.
So, in addition to the foreign AI applications of Midjourney, Ideogram, and stable diffusion 3, what are the domestic AI applications of Wenshengtu that may replace them?
I have compiled a list of domestic AI mapping applications that may be able to replace Midjourney for your understanding and reference when needed.
Baidu -Literary Mind
Alibaba-Tongyi Wanxiang
Beautiful pictures-WHEE
SenseTime-Seconds
Vega AI
Tiamat AI
Let’s take a look at their detailed introduction below.
Baidu - Wenxinyige
Official Website
Wenxinyige official website address:https://www.1ai.net/8595.html
Introduction
Wenxin Yige uses Baidu Wenxin ERNIE-ViLG 2.0 as the large model base for AI painting. The model adopts knowledge enhancement algorithm and hybrid expert diffusion model modeling, learns massive high-quality text and image data, comprehensively improves the image generation quality and semantic consistency, and achieves world-leading results in the field of text and image generation.
team
The team behind Wenxin Yige is Baidu's AI research and development team, which is responsible for the platform's technological innovation and product development.
Function
The functions of Wenxinyige include but are not limited to:
- Text to Image: Users can generate images by entering descriptive text.
- Multi-style adaptation: Supports image generation in multiple artistic styles.
- Smart Recommendations: Intelligently recommend the style and content of the generated image based on the description entered by the user.
- From pictures to images: After the user uploads the picture, AI can generate a new image based on the content of the picture.
- Secondary Edit: Perform secondary editing and adjustment on the generated image.
Free or paid usage instructions
It takes 8 batteries to generate one image.
Electricity is divided into paid electricity and free electricity. The differences between the two are as follows:
A power pack of 80 charges costs 9.9 yuan and is valid for one year, which is equivalent to generating a picture at a unit price of 0.99 yuan.
Alibaba-Tongyi Wanxiang
Official Website
Tongyi Wanxiang official website address:https://www.1ai.net/1876.html
Introduction
Tongyi Wanxiang is an AI painting creation model based on the self-developed Composer composition generation framework, which can generate images of different styles that conform to semantic descriptions based on the text content entered by the user. It performs well in semantic understanding and image generation, and supports image generation in multiple styles, such as watercolor, flat illustration, two-dimensional, oil painting, Chinese painting, 3D cartoon and sketch, etc.
team
The team behind Tongyi Wanxiang is Alibaba Cloud Intelligence Group's CTO Zhou Jingren and his team, who are responsible for the research and development and innovation of the model.
Function
The functions of Tongyi Wanxiang include:
- Text to Image: Generate an image based on text content.
- Partial repaint: Redraw and generate the unsatisfactory parts of the image.
- From pictures to images: Generate a new image based on the uploaded reference image.
- Graffiti painting: Generate exquisite graffiti paintings by adding hand-drawn content and text description.
- Photo Gallery:Upload 2-4 personal photos to get an exclusive photo album.
- Word Art:Convert text into beautiful Word Art.
Free or paid usage instructions
Under normal circumstances, each time you generate an image, you need 1 inspiration value. (Depending on the selected ability, the inspiration value consumed may be different)
You will be given 20 inspiration points for your first login, and 3 inspiration points will be automatically given for your daily login.
Purchase is currently unavailable.
The specific rules for issuing inspiration points (screenshot on April 20, 2024) are as follows:
Beautiful pictures-WHEE
Official Website
WHEE official website address:https://www.1ai.net/2180.html
Introduction
Through its core AI technology, WHEE allows users to input text descriptions or upload pictures to quickly generate artistic images and paintings. It not only includes basic functions such as text-to-image, picture-to-image, line drawing coloring, and graffiti-to-image, but also provides advanced functions such as AI photo editing to meet the diverse needs of professional designers in visual creation.
MiracleVision, the AI model used by WHEE, was independently developed by Meitu. The model learns the relationship between descriptive words and images by analyzing a large amount of graphic and text data, thereby realizing the ability to convert natural language descriptions into visual content.
team
The team behind WHEE is Meitu, which has a professional technical team responsible for R&D and innovation. They are committed to applying the latest AI technology to the field of visual creation.
Function
The main functions of WHEE include but are not limited to:
- Vincent Figure: Generate corresponding images based on the user's text description.
- Figure 1: Generate new artwork based on uploaded images and attached descriptions.
- AI Image Enlargement: Automatically expand the image background.
- Graffiti: Transform simple hand-drawn sketches into sophisticated images.
- AI Photo Editing: Provides image editing functions such as high-definition enlargement and style transfer.
Free or paid usage instructions
Each image generation costs 8 USD beans.
New users will be given 80 US beans when they log in, which is equivalent to 10 opportunities to generate pictures.
The recharge of US beans is 1 yuan = 10 US beans, which is valid forever. This is equivalent to the unit price of generating a picture at 0.8 yuan.
360-360 Smart Drawing
Official Website
360 Zhihui official website address:https://www.1ai.net/8597.html
Introduction
360 Zhihui aims to help users turn their imagination into reality, providing functions including text-generated drawing, image-generated drawing, graffiti-generated drawing, partial redrawing, custom model training, etc. The tool uses AI technology, supports drawing prompts in both Chinese and English, can quickly generate pictures, and provides a rich model library and stylization options.
team
The team behind 360 Zhihui is Qihoo 360 Technology, a team with deep accumulation in the fields of Internet security and artificial intelligence technology.
Function
The main functions of 360 Zhihui include:
- Text to image: Supports bilingual AI drawing prompts in Chinese and English, and generates pictures based on descriptions.
- Image Generate Image: Users can upload pictures, and AI will generate paintings in other styles based on the uploaded pictures.
- Graffiti: Simply draw a few strokes and generate in real time.
- Partial repaint: Change whatever you want.
- Custom Model:Users can upload personal image materials and train their own style models.
Free or paid usage instructions
The official has not specified the number of free trials per day. (Guotou has 2 trial opportunities when he logs into 360 Zhihui on April 20, 2024)
You need to recharge your membership to use the 360 Zhihui platform, which costs 25 yuan per month. (The upper limit of the number of times a member can use the platform is not specified.)
SenseTime-Miaohua
Official Website
Miaohua official website address:https://www.1ai.net/8599.html
Introduction
Miaohua is an AI painting creation platform launched by SenseTime. It is based on SenseTime's self-developed AIGC large model and provides a series of convenient LoRA training capabilities. The platform aims to provide creators with convenient and complete content production and creation tools, allowing users to generate artistic works through simple Chinese prompts, supporting a variety of styles such as photography, cute, exquisite, cyberpunk, movies, etc.
team
The team behind Miaohua is SenseTime, a company with deep accumulation in the field of artificial intelligence, strong R&D strength and technological innovation capabilities.
Function
The main functions of Miaohua include:
- Vincent Figure：The user enters the text prompt and AI can generate the corresponding image.
- Figure 1: Based on the reference images provided by the user, AI generates new paintings.
- LoRA model training：Users can upload their own images and combine SenseTime’s self-developed painting model or open source model to train a customized LoRA model to generate personalized content.
- Aesthetic Guidance System: The new version uses SenseTime’s self-developed Artist v0.3.5 model, which has a built-in aesthetic guidance system to enhance the artistry and visual quality of the images.
Free or paid usage instructions
It is currently free to use. The picture below is a screenshot of Guotou logging into Miaohua on April 20, 2024.
The number of times you can create photos every day is 10 times/day. If you invite friends, you can get a 3-day unlimited card.
Tiamat AI
Official Website
Tiamat AI official website address:https://www.1ai.net/8602.html
Introduction
Tiamat AI is a leading online AI painting platform in China. It was founded in 2021 and is legally owned and operated by Queshi Digital Technology (Shanghai) Co., Ltd. and its affiliates. It uses advanced AI technology to allow users to generate artistic images through text descriptions. Developed by a startup team incubated with the support of ShanghaiTech University, the platform aims to provide AI services in the form of creative production, advertising, digital collections and other scenarios for individual users and corporate customers. Tiamat AI is known for its self-developed MorpherVLM model, which is the first cross-modal generation model in China with a capacity of nearly 10 billion based on the concept fusion paradigm.
The name "Tiamat" comes from Tiamat, the goddess of creation in ancient Babylonian mythology, and was named by voting by early core community users.
team
The Tiamat team was incubated by ShanghaiTech University and is composed of senior industry practitioners. Team members have backgrounds in fine arts and computer science and technology. Founder and CEO Qinggan has ten years of experience in fine arts, and later switched to computer science and technology, leading the team to focus on the research and development and application of AI generation technology.
Function
Key features of Tiamat AI include:
- Text to Image：Users input text descriptions and AI can generate corresponding image works.
- From pictures to images:Upload a reference image and describe the image to generate a new image.
- Tons of templates and models: Provides a variety of AI painting templates and models, supporting image generation in different styles.
- AI Search and Gallery: Provide users with AI search functions and image galleries to facilitate the acquisition of creative inspiration and storage of works.
Free or paid usage instructions
Register and get 20 standard image generation times.
After using up the times, you need to purchase spirit stones or recharge VIP to get more picture generation times.
1 yuan = 60 Lingshi, 20 Lingshi = generate picture 10 times, so 1 yuan = generate picture 30 times, which is equivalent to 0.03 yuan per time to generate picture.
Vega AI
Official Website
Vega AI official website address:https://www.1ai.net/8604.html
Introduction
Vega AI is an AI painting tool launched by RightBrain AI. It uses deep learning technology to generate artistic images based on simple sketches or descriptions input by users. Vega AI aims to lower the technical threshold of painting, allowing users without painting foundation to create high-quality works of art. The platform supports a variety of painting styles and modes, is easy to operate, and is suitable for users of all levels.
Founded in September 2022, Beijing RightBrain AI is a startup company focusing on the research and development of AI image and video generation. It is committed to applying AIGC technology to the fields of images and videos, empowering creation and making imagination concrete.
team
All the members of Right Brain's founding team graduated from Peking University. They are core members of NUWA, Zhiyuan Wudao and other multimodal pre-trained large models. They have the background of top domestic AI research in Microsoft Research Asia, SenseTime Research Institute, Zhiyuan Research Institute, etc. They have the most cutting-edge research and implementation experience in image and video generation in the industry, and several results have been published in top conferences such as NeurIPS, ECCV, and ACM MM. At present, it has completed the seed round of Qiji Chuangtan founded by Dr. Lu Qi, as well as tens of millions of angel rounds of financing from first-tier VCs such as Lightspeed China. As one of the earliest researchers engaged in AI visual generation in China, Right Brain Technology has rich experience in large model research. It is committed to reshaping the AIGC (Artificial Intelligence Generated Content) creation ecology and continuously unleashing creative imagination.
Function
Key features of Vega AI include:
- Vincent Figure: Users enter descriptive text and AI can generate corresponding images.
- Figure 1: AI generates new pictures based on pictures uploaded by users.
- Vincent video:Users enter descriptive text and AI can generate corresponding videos.
- Figure video:Based on the pictures uploaded by users, AI can generate corresponding videos.
- Style training: Supports online rapid training of AI painting models, and users can freely customize the generated effects.
Free or paid usage instructions
The basic functions of Vega AI are free to use, and a trial period is provided. Details are as follows:
Image generation
- HD: 200 free builds per user per day
- Ultra HD: 25 free times per account
PS: 200 HD pictures per day, basically you can never use them all.
Video Generation
- Each account has 5 free trials
When the number of uses is used up, you can recharge your membership to get new times.
4.99 yuan = 150 ultra-clear images, which is equivalent to generating ultra-clear images at 0.03 yuan per time.
4.99 yuan = video generation 10 times, which is equivalent to 0.5 yuan per video generation.