GuShiio.com Wen| Tian Wei AI
Early in the morning, I turned on my computer and clicked on it habituallyAIdialog tool。”Help me write a report on urban planning, which requires data analysis and case comparison. “A few seconds later, a clearly structured report appeared on the screen. Afterwards, I usedAI drawing toolsSeveral matching pictures were generated and AI was usedaudio toolThe generation tool recorded a voice explanation, and in less than half an hour, a preliminary project plan was born.
Looking back two years ago, the same job might have required me to spend a whole day reviewing materials, writing content, looking for pictures, and making slide shows. Now, AI has quietly changed the way we work, just as the Internet and smartphones did, reshaping our lives in an irreversible way.
In 2025, the AI industry has ushered in an eye-catching development moment. At the beginning of the year, the popularity of DeepSeek allowed domestic users to truly experience the powerful capabilities of the big language model, setting off a wave of AI learning craze. Looking back on the development process of AI in the past three years, a relatively mature application pattern has been formed.
It mainly focuses on AI dialogue, AI writing, AI painting, AI video generation, AI audio generation, AI programming, AI search and AI music.
In this booming market, various companies are actively deploying and competing for their own place. Today, leading products in various segments have gradually become clear, and their growth rate has slowed down.
So as an AI enthusiast, I compiled this guide, hoping to help friends who are new to AI avoid detours and find the most suitable tool for them.
It should be noted that most AI products actually have a lot of overlap in functions. When classifying, I mainly consider their core advantages and specialties. For example, although DeepSeek is also a good AI conversation tool, I think it performs the best in writing, so I classified it into the category of AI writing tools.
This article is 6,000 words and is quite detailed. I will analyze all types as clearly as possibleAIAdvantages and disadvantages of the product and application scenarios.
I suggest you collect it first and study it slowly when you have time. You are also welcome to like it, collect and share it. Every encouragement you give is the motivation for me to continue to create.
At the same time, some tools require magic to use, so this article does not classify them. If necessary, it is recommended to understand them yourself.
To show sincerity, please attach a version that is too long to read:
1. AI Dialogue
Features: The most comprehensive function, the most abundant ecology, and the most widely usedAIDialogue platforms are almost omnipotent.
? Link:https://chatgpt.com
Currently the most comprehensive AI product in terms of multimodal capabilities. The product experience is very good. The voice chat is smooth and natural, and the response is quick. When making video calls with AI through the camera, it can directly identify real objects, and even simulate human emotional changes.
I used it myself and did perform best among companionship AI. Both the response speed and the feedback on the conversation content are in place, giving people the feeling of communicating with real people.
However, as the oldest AI tool, to be honest, it is a bit cheap and old, and there has been some stagnation in product innovation in the past two years. Although the same company’s Dall-E3 and Sora have technical highlights, they are not practical enough and are surpassed by competitors. The recent launch of ChatGPT 4.5 also failed to bring many surprises, with mixed reviews on social media.
In addition, CEO Sam Ultraman’s business orientation has also caused a lot of controversy. Although OpenAI maintains a certain technological lead due to its first-mover advantage, there is still room for improvement in product practicality and user experience.
Characteristics: The best understanding of the Chinese context.groundHe is passionate and professional, has a full product experience, and is a close partner for domestic users.
? Link:https://www.doubao.com
In the past two years, the development of domestic AI has been limited by computing power bottlenecks, and its basic performance is generally weak. Until the emergence of DeepSeek, it was difficult to truly become a productivity tool.
However, ByteDance makes up for the lack of model capabilities with its powerful product polishing capabilities. As the Internet company with the strongest product power at present, they are indeed outstanding in terms of user experience and have successfully developed many practical and interesting chat robots.
Daily English practice, tarot card interpretation and other functions are provided by Doubao AI, and the user experience is quite smooth.It is highly recommended to install one for the elderly at home. If you have nothing to say, you can relieve boredom without being afraid of the elderly being cheated by electricity.
Features: Google’sAIThe product integrates Google’s powerful search capabilities.
? Link:https://gemini.google.com
As Google’s AI product, Gemini combines Google’s powerful search capabilities and AI dialogue capabilities to create a unique experience.
The biggest highlight of Gemini is its seamless integration with Google Search. When I need the latest information or specific data, Gemini is able to search and integrate information in real time, an advantage that other dialogue models cannot achieve.
2. AI writing
Features: The country’s strongest open source model and its creative writing capabilities are explosive. They are an important weapon for Chinese people to achieve “AI freedom.”
? Link:https://chat.deepseek.com
The strongest open source model in China has set off the AI craze in 2025. It is also the most powerful model that Chinese people can easily use. Now everyone has access to DeepSeek, allowing us to basically realize AI freedom.
In my actual use, DeepSeek has performed well in creativity, writing, etc., and has reached the ability to truly serve as a big model of productivity.
It is more than enough to write small red book copywriting, public account articles, etc. Students engaged in the creative industry can also use it to inspire inspiration.
However, the shortcomings are also obvious. The output content is sometimes too exaggerated, which leads to reduced usability and the illusion rate remains high.
However, the latest news is that DeepSeek R2 will be released soon, and I hope this problem can be improved.
Characteristics: High quality writing and code writing, super understanding, and a right-hand assistant for professional writers and programmers.
? Link:https://claude.ai
It is recognized in the industry as the largest model with the strongest writing and coding capabilities. It can understand user needs very well and output language with a strong sense of real life. The recently released Claude 3.7 is the first hybrid model, and its capabilities are further enhanced.
However, the boss’s situation is not big. When DeepSeek came out before, he also posted a message maliciously speculating on magic formulas. What’s more regrettable is that the China account was severely blocked. I can only say that I love and hate it.
Characteristics: Full sense of humor, daring to speak, and entertaining peopleAI。
? Link:https://grok.com
Musk’s money-using capabilities and the latest model piled up with 200,000 GPUs, I can only say that it’s great to have money. This AI is very interesting. It talks like a funny person and is very abstract. When I use it, I always feel like Musk is replying to me backstage.
The ability to understand Chinese is also good, but writing social media content is very suitable. It can catch hot spots on the Internet, play tricks well, and the content written is both interesting and resonates.
You can use it to blow water every day. After all, many of the language materials used by Grok are on X. X is a platform that compares with no taboos and can talk about anything.
The picture below is the answer given by Musk, who asked it to evaluate his boss, without anyone setting it up. I am so happy.
3. AI painting
Features: Designed specifically for Chinese users, has a deep understanding of Oriental aesthetics, and is a right-hand assistant to domestic creators.
? Link:https://jimeng.jianying.com
The AI painting product launched by Byte, like the previous bean bags, has a full user experience and a very smooth operation. Moreover, the support for Chinese painting prompts is very good. You don’t have to rack your brains to think about English descriptions, and you can draw amazing works directly in Chinese.
I used it to generate several China style illustrations last week, and the effect was amazing! That kind of ink charm and oriental artistic conception are well balanced and do not seem to be generated by a machine. Especially when dealing with China elements such as Han costumes and ancient buildings, Dream knows much more than foreign models, and the details are handled in place.
It also supports the direct generation of posters with Chinese characters. You should know that most of the Chinese generated by AI paintings are garbled. I used AI direct solar term posters before, and the effect was still amazing.
Features: Launched quicklyAIPainting tools are simple and easy to use.
? Link:https://klingai.kuaishou.com
The fast-handed AI painting tool has high quality picture generation and also supports Chinese prompt words. It is a high-quality picture generation AI tool that ordinary people can easily use.
AI clothes changing can change models ‘clothes with one click. It is highly recommended for e-commerce partners to try it.
Features: It does not rely on text descriptions, but freely combines multiple pictures by uploading them, which is simple and intuitive.
? Link:https://labs.google/fx/zh/tools/whisk
The AI painting product launched by Google is essentially different from traditional AI painting tools. It no longer relies mainly on text prompts, but allows users to upload pictures as the main input method to achieve a more intuitive artistic creation process.
We can take a photo of our own cat as the “subject”, choose an oil painting as the “style”, add a scene in an autumn park, and in a few seconds we can get an art painting of a cat gentleman wearing a formal suit in an autumn park!
It is very playable. I used this tool to create a large number of very interesting pictures new year ago, integrating my favorite IP images into real scenes.
Features: Focus on commercial design scenarios, from icons toUI, a professional tool to solve design needs in a one-stop manner.
? Link:https://www.recraft.ai
There are many built-in styles, and you can easily draw the pictures you want. At present, all my own public account pictures are drawn by Recraft.ai, and the effect is very good.
Features: Powerful technology, exquisite details, and high integration. It is an ideal choice for pursuing high-quality image generation.
? Link:https://labs.google/fx/zh/tools/image-fx
I used the most realistic Wensheng AI painting tool I have ever used, and I also understood the prompt words well.
Picture 1 below: I asked him to generate a picture of a robot illusion. Other products did not understand the illusion properly. Only ImageFX achieved the effect in my mind.
The spider effect generated by the second picture is so realistic that friends thought it was a real shot! The treatment of light, shadow, picture details and other aspects is particularly in place.
6.Midjourney
Characteristics: ProfessionalAIPainting, strong ability and complex operation.
? Link:https://www.midjourney.com
The earliest AI image generation tool that came out of the circle is professional and powerful. However, the threshold for use is high and the operation is slightly complicated. Moreover, the price is not cheap and is more suitable for professional designers.
4. AI video
Features: The domestic leader in the field of video generation launched by Quick Hand, with extreme picture quality and smooth movements.
? Link:https://klingai.kuaishou.com
Keling is DeepSeek in the field of video generation. In my memory, it is also the first domestic AI tool to catch up with top international similar products.
The powerful video generation capabilities have made the international version of Kelin popular by foreigners even after it was launched. The generated video is of extremely high quality, without confusion, and has a good understanding of the actions.
To be honest, before the fast hand came out of Kokao, I was a little biased against the high and low level of the old iron, but since the Kokao was used recently, I can only say that the fast hand is amazing *.
Characteristics: Top Chinese people foundedAIThe video platform has excellent ability to integrate dynamic elements.
? Link:https://pika.art
It is also an AI video company founded by Chinese people, and its overall video generation quality is first-class.
A new feature was recently launched that allows you to add other dynamic elements to videos. The left side of the picture below is the original video, and the right side is the AI-generated video. You can see that the generated video is very effective and highly realistic.
If you want to play, or if you need it, you can try Pika.
Features:AIA pioneer in the video field, the price is high but the professional quality is irreplaceable.
? Link:https://runwayml.com
OpenAi in the AI video field, when Gen-2 was first released, was the day in the video generation field. Just enter a few descriptions or upload a picture, and it can generate a smooth and natural video for you. It’s not that kind of rough animation, but a realistic picture that even professionals would be amazed at.
Although the AI video field is full of stars nowadays, Runway is still not old. The only unfriendly thing is that the price is too expensive and it needs to be copied. If it is too troublesome, I recommend everyone to use Keri directly.
Features: Domestic videolarge modelThe people-friendly choice, the Chinese interface is friendly and easy to use, and the price is extremely high
? Link:https://www.vidu.com/zh
It is also a first-class domestic video model, and the effect is good. The user interface is also very friendly to Chinese people and supports Chinese prompt words, which is very easy to use.
And the main one is cheap. If you are just interested in AI videos, you can try playing Vidu.
5. AI audio
Features: MINIMAX’s sound generation artifact, natural reading, accurate and realistic cloning, and a dubbing tool for self-media creators.
? Link:https://hailuoai.video
MINIMAX products are covered in everything from pictures to videos to sounds, and the effects are all good.
I personally like to use the sound generation function of Conch AI best. The text reading is clear and can be automatically frustrated, with a strong sense of reality.
In particular, the sound cloning function can directly generate audio based on my tone. Currently, the audio I play from my media is made by Conch. I will insert a paragraph below to show you the effect.
It has always been free before, but recently it has been charged, but check-in time can be given away, and it is basically enough for daily use of white prostitutes. What kind of cyber Bodhisattva is MINIMAX?
6. AI programming
Characteristics: Cursor has the most professional ability and the most complex operation;Windsurf comes second;Trae comes second.
? Link:
https://www.cursor.com
https://codeium.com/windsurf
https://www.trae.ai
I am not a programmer myself and have little research on AI programming tools, so I don’t pretend to be a big-tailed wolf and recommend it to everyone.
I have used Cursor, Windsurf, and Trae, and I have used them to make complete products.
In terms of my simple experience, Curosr is more professional in AI programming, but the barrier to use is higher.
Windsurf is more suitable for newbies and is easier to operate.
7. AI search
Features:AIThe originator of search tools.
? Link:perplexity.ai
The earliest AI search tool combines AI and search, greatly improving the accuracy of information.
It is very suitable as a supplement to search engines such as Google, and is also a tool for scientific researchers such as content creators to retrieve data.
The shortcomings are, after all, foreign tools and have not been optimized for China. At present, domestic nano-AI search and secret tower search have been developed, and there are better products.
Features: Zhou Hongyi launchedAISearch tool, comes with many AI tools, office tools
? Link:https://www.n.cn
The nano-AI search launched by Zhou Hongyi took advantage of the DeepSeek wave to quickly launch a free full-blooded version of the R1 reasoning model. The download volume has increased rapidly, and the daily activity has also increased rapidly.
The user experience is not bad, and there are necessary functions. The search results basically meet my requirements, but sometimes factual errors will appear. It has a lot to do with the capabilities of the DeepSeek R1 model itself.
In addition to search, there is also a multi-model collaboration function that allows AI to play different roles to check for gaps in answers. There are also some practical AI tools that help us improve our daily efficiency, which are generally very suitable for migrant workers.
Characteristics: Focus on scientific research and learningAISearch tools, you can build your own knowledge base and industry analysis reports, which are necessary for work, study and home.
? Link:https://metaso.cn
It is similar in function to nano-AI search, and is also connected to DeepSeek r1. But the overall tool is more like a
In addition to AI search, the learning tool also comes with a document search function. Relevant documents can only be converted from English to Chinese. It is very convenient. Ma no longer has to worry about my poor English and not being able to understand foreign languages.
At the same time, you can also upload the required documents to the secret tower to build your own knowledge base. It can only be said that the secret tower search is a necessity for academic research, work study, and home.“”
8. AI music
Features: AIA pioneer in the field of music creation, supports diverse musical styles and is highly professional.
? Link:https://suno.com
The earliest and most popular AI song generation tool is currently the best to use, supporting the generation of a large number of music styles. I have used it to make a lot of lofi music myself, and the animation generated by AI video is very good.
At present, there are already a large number of lofi content generated by Suno on YouTube for AI music +AI videos, which is a way to make money. The video below is a pure music song I tried to produce before.
Features: Popularization launched by ByteDanceAIMusic tools are easy to use with low barriers.
? Link:https://www.haimian.com
Another is the AI product launched by Byte. From this series of products, we can see that Byte has restarted the App factory, and Zhang Yiming is bound to win the AI mountain.
Overall, Sponge Music, like other AI products of Bytehome, has a low threshold for use, and the music generated is okay and basically enough. If you like music, you can play more.
Features: Produced by Google, simple to operate and high quality generation
? Link:https://labs.google/fx/zh/tools/music-fx
Google’s new AI music product is simple to operate. I often use it to generate lofi music. There is not much research on other usages.
It can be seen that Google has begun to make efforts in the field of AI, and the momentum is very strong, with a strong intention of catching up from behind.
Google is now part of almost every field of AI. It can’t be better for our ordinary users. After all, the products of big factories are not only of high quality, but also basically free of money. Hehe.
Standing today in 2025, we are at the center of this change. AI tools are constantly updated iteratively, and each update expands the boundaries of possibilities.
As witnesses and participants of this era, we must not only maintain an open attitude towards new technologies, but also maintain the ability to think independently; we must not only make good use of AI to improve efficiency, but also enhance our irreplaceability through continuous learning.
In the future, the most valuable people are not those who will use specific AI tools, but those who can integrate multiple tools, think across borders, and solve problems creatively.
Just like every technological revolution in history, the development of AI will eventually allow us to rethink the unique value of mankind and discover new possibilities for ourselves in this thinking process.
That’s all for this article’s sharing of AI tools. Nowadays, AI tools are blooming everywhere. This article is written based on my own experience, and there must be mistakes or incomprehensiveness.
If there are any mistakes in this article, readers are welcome to correct them. If you think I wrote well, you are welcome to give me a reward, like it, and forward it. Every encouragement you give is my greatest motivation to continue to create.
