Recently, ChatGPT has upgraded the GPT-4o model and Dall-E3. Now, it can easily create its own LINE stickers or interesting four-frame comics through ChatGPT’s image generation function. It can even customize Ghibli styles, Snoopy comics or various meme pictures with generated pictures. The following “Mr. Crazy” will introduce how to use ChatGPT-4o to generate LINE stickers, style comics and Ghibli style pictures respectively!
ChatGPT generates LINE map method
Step 1. Upload the original photo to be made into a sticker
After opening the ChatGPT chat room, upload a photo as a reference, which can be a selfie or a pet photo, and confirm whether the ChatGPT-4o model status is presented at the top.
Reminder: Currently, ChatGPT-4o’s drawing generation function requires a paid member to use it. This function cannot be used if it is a free account.
Step 2. Enter the generated map style and instructions
Enter the following command to generate LINE map in the ChatGPT dialog box, and press to send after the input is completed.
Use the (animal or human) in the photo to create x (according to personal needs) Q-patterns, and add the following text for each photo (GPT can be required to play without input in a custom situation) etc.
Step 3: Adjust the optimization details
ChatGPT will generate a map based on the instructions just mentioned and Q-versioned according to the original picture. It will appear after about 1 minute. If you are not satisfied with the generated result, you can further ask GPT to adjust the color, movement, expression or text.
Dog stickers generated by ChatGPT.
It is also OK to use your big head photo to make Q-version stickers. ChatGPT will automatically depict the face shape and hairstyle characteristics of the photo. Even if you don’t add text, there are also many different expression patterns.
Q-version portrait map generated by ChatGPT
Step 4: Split the picture and remove the back
If you feel that the map generated by ChatGPT is very satisfied and want to download it yourself, or plan to upload it to LINE for sale, you can issue “Please divide and memorize each image” to GPT, and GPT will automatically divide and memorize it for you, and then there will be independent LINE maps that can be used or uploaded to LINE Creators Market for itself.
But be aware that GPT is not very good in cutting technology, so it is recommended to manually cut and modify it yourself, which will be more exquisite. If you want to upload it to LINE map for sale, the size of each image is limited to 370 pixels wide × 320 pixels high.
ChatGPT generates four-frame comics
It is no problem to generate four-frame comics through the built-in drawing of ChatGPT-4o. You must first come up with the theme and content of the comics, such as the following example:
Pure contextual comic style
You can use ChatGPT to require “generate a complete Japanese commercial color illustration style with “theme situation””, and then GPT will directly draw a comic you want to present for you.
Text-type comic style
If you want to add text words to each comic, you can also use the following ChatGPT 4o to generate comic instructions. You can replace the situation and text you want to change by yourself, and you can quickly generate it.
Please use four-frame comics to present the situation of “sorrow people often scolding Apple”, which should include “selling so expensive, Apple fans are plates”, “Buying iPhones is just showing off”, “You will only change colors and charging holes in a year”, and “A charging cable is so expensive”.
Please use four-frame comics to present the retail trading situation, which should show “chasing highs and killing lows”, “knowing the cards to buy stocks”, “missing buying points, hesitating”, and “trapping (long-term loss holding)”
Of course, it can also be integrated into the Japanese black and white comic style to generate, and the following methods can be used to issue instructions:
Generate a picture that presents the favorite message of malicious people in the Japanese comic style, which will show you “Selling expensive apple powder is a plate”, “Buying an iPhone is just a show”, “I will only change the color and thicker electric holes in a year”, and “A charging cable is so expensive to sell.”
Integrate cartoon characters comic style
Of course, you can also customize drawings according to a cartoon character, such as the classic stock market meme plot that can design retail trading situations presented in the style of “Sailor Moon”. Each frame is accompanied by a picture description and text suggestion lines. However, the current GPT is not very stable in Chinese fonts, and some fonts will be broken or messy, but you can try to generate them a few more times. Some churches are normal.
Please use the four-frame comic to present the retail trading situation of Sailor Moon, which shows “chasing highs and killing lows”, “knowing the cards to buy stocks”, “missing buying points, hesitating”, and “trapping (long-term loss holding)”
ChatGPT generated Sailor Moon style comic pictures
ChatGPT generated Sailor Moon style comic pictures
ChatGPT Generates Ghibli Style Picture Skills
If you plan to convert your personal photos directly into Ghibli style, just upload the pictures you want to make to ChatGPT and issue the command “Generate Ghibli style pictures” and GPT can immediately depict the entire picture in Ghibli comic style, with a super high texture and super realistic!
Chat-GPT 4o genre prompt words (description direction)
In addition to using ChatGPT 4o to generate Ghibli-style pictures, there are actually many picture styles that can also be used together. Below we will sort out the current ChatGPT generated picture style types.
| Style Type | describe |
|---|---|
| Q-version image | Super cute, with a big head and a small body, like a cartoon character |
| Picasso | Don’t look like a real person, with strange lines but distinctive, and a strong sense of art |
| American comics | Like superhero comics, the characters are very distinctive, the actions are exaggerated and dramatic. |
| Bauhaus | It looks simple, practical and has a sense of design, like modern architecture or furniture |
| Graffiti style | Free and casual, like graffiti on the street walls, it feels like youthful and energetic |
| Ghibli style | A bit like Hayao Miyazaki’s animation, warm and a little fantastic |
| Silhouette | Just look at the outline of the object, just like a shadow, simple and tangible |
| ASCII art | Pictures arranged in English letters or symbols are much like art pictures on computer screens before |
| 3D cartoon | It has three-dimensional sense, looks like animated characters, very vivid |
| African Futurism | Combining African culture with a sense of technology makes it look cool and creative |
| Mondrian | A picture made of squares and several basic colors looks simple but modern |
| Mu Xia | The picture is beautiful and gorgeous, the characters are soft, like an old poster |
| Broken effect | The image looks like it is broken, a bit handsome and a bit messy visual effect |
| Bohemian style | There are many colors and patterns, and they look free and romantic |
Summarize
Through ChatGPT-4o, the powerful generation of pictures allows creativity to no longer be limited by technical thresholds. Whether it is personalized LINE stickers or interesting four-frame comics, or even converting pictures into various styles such as Ghibli, you can quickly complete the exclusive works with simple instructions and some adjustments. Come and try it together!
More practical tips:
