Ji Meng 2.1 is incredibly entertaining. Last night, I was captivated by the grayscale images and models of Ji Meng 2.1, and since then I have been exploring various prompts. I manage to exhaust the daily bonus points every day, and today’s 100 points have already been used up. I simply can’t stop; it’s just too much fun. After exhausting my Ji Meng points, I went to Dou Bao and Kou Zi to get more 2.
1 image models… Yes, if Ji Meng hadn’t been updated to 2.1, everyone could experience it at Dou Bao. Moreover, Kou Zi also supports 2.1 image models (text-to-image and image-to-image). I shared 20 cases where Ji Meng 2.1 can generate text, but I didn’t expect it to produce such impressive text with so many ways to play. Below are some good cases and methods I’ve run on Ji Meng over the past two days, along with the prompts, which I’m sharing with everyone.1. Movie poster, fantasy style, featuring a beautiful woman riding a giant cat. Bottom text: ‘The Adventure of Little Meng Cat Kingdom,’ releasing for the 2025 Spring Festival.
2. Live-action shooting, imprinting the words ‘Wo Yin AI’ on a building. 3. Live-action shooting, internet company with a tech feel, with the words ‘Wo Yin AI’ at the front desk. The lighting in this image is really impressive, even the reflection of the words on the desk is well done. Yes, it’s a real shot, and some of the lights in the logo are broken; I’ve already asked the company’s front desk to fix it. 4. Product display, imprinting the words ‘Wo Yin AI’ on a graphics card. 5. Product display, imprinting the words ‘Wo Yin AI’ on a pack of tissues. 6. Product display, imprinting the words ‘Wo Yin AI’ on a bottle of mineral water. 7. Product display, marking the words ‘Wo Yin AI’ on Bluetooth earphones. 8. In the vast autumn of the desert, the grass withers, and the few soldiers fight at the setting sun in the lonely city. Always under grace, I often underestimate the enemy, and the effort is exhausted, but the siege is not lifted. Right vertical title: ‘Song of the Swallow’. 9. Heroes go to Liangshan, and romantic stories are passed down. The court frequently raids and washes, but the stockade is at ease. The seats of the evil stars and the heavenly stars are noisy with flags and drums. A thousand years of wind and rain have passed, and the setting sun shines on the book. Right upper vertical title: ‘Heroes of Liangshan’. 10. There is a country overseas, called Ao Lai Country. The country is close to the sea, and there is a famous mountain in the sea, called Hua Guo Mountain. This mountain is the origin of the ten continents and the coming dragon of the three islands. It has been established since the beginning of clear and turbid, and it has been formed after the judgment of Hongmeng. What a good mountain! Title: ‘Hua Guo Mountain’. 11. A distant view, aerial photography, western Sichuan snow-capped mountains, layers upon layers, and the mountain groups are scattered. White text ‘Western Sichuan is really beautiful’ blends with the snow-capped mountains, with the lower half of the text transparent and hidden behind the mountains. Friends who like to brush travel videos must have seen this text effect, with super large place names standing between the peaks, and some text hidden behind the peaks. Now I use Ji Meng AI to reproduce this effect with just one sentence.If you have your own landscape pictures, you can also send the images to the AI. Just modify the prompt words I mentioned above a little bit, and let the AI generate special effect text. 12. Year of the Snake poster, Chinese red, Spring Festival, jubilant, with good implications, handwritten snake-shaped text “Hello 2025”. 13. Year of the Snake poster, New Year, snake, cute, festive, with good implications, light green, text “The Auspicious Snake Brings Good Fortune”. 14. Movie poster, ancient martial arts, the back view of a female knight, the reflection of the figure on the lake surface, Chinese-style photography, sunset glow, red moon, cyan-orange color tone, Oriental aesthetics, high-class movie feel. The vertical text on the right is “The Gallant Swordsman”. 15. A woman lives inside a glowing shell. The picture is simple and glittering. The title is “The Woman of the Conch”. 16. NFT art, pixel style, city night view, neon lights, the seaside road at night, glitch style, particle special effects. A car is driving on the seaside road, with pixel squares of different sizes, stream of consciousness. The title is “2046”. 17. Science fiction movie poster, pixelated and colorful artificial intelligence collages, sense of speed, blurring, high-class feeling, strong contrast between cold and warm colors, digital glitch aesthetics, subversion and reshaping of the existing system. The top title is “The Wandering Earth”, text “Production: Leng Yi”. 18. Mountains in the mist, high-definition wallpaper, dreamy, minimalist, aesthetically healing, Iceland, a red mist horizontal line, a dreamy artistic conception with a sense of story. The text at the bottom is “I Have Been Here”. 19. A cute and funny anime-style panda, cute anime illustration style, facial features, rich expressions. A four-grid picture shows four expressions of sadness, anger, thinking, and shyness respectively. Each grid has the text “Sadness”, “Anger”, “Thinking”, and “Shyness”. Familiar friends must know that this prompt word is modified based on the previous meme prompt word. At that time, the AI could not generate text. Unexpectedly, now it can directly generate text. It’s really convenient. 20. Dreamcore aesthetics, liminal space, glitch art, vaporwave aesthetics, racing car, surrealism. The text centered at the bottom is “The Past of the Future”. After reading these cases, what do you think? Are you quite shocked? As a layman in design, I was quite shocked. But not long after, I got out of the shock and then started playing crazily. It’s really fun. It can definitely help me a lot. Whether it’s posters, product pictures, brochures, memes, or article cover pictures, I can now use it to directly generate pictures. The most crucial thing is that it is very intelligent. You don’t need to emphasize xx font, xx color, xx position, xx font size in the prompt word. You just need to say what text to put, and it will design the text and layout according to the picture style by itself. The quality of the generated pictures is very high.
Of course, if you are a designer, the benefits of efficiency enhancement are even more significant. The model is trained based on the expertise and knowledge of designers, ranging from aesthetics, visual perception to composition. If you can articulate these aspects precisely through language, AI can replicate your skills. Thus, the Dream 2.1 image model is actually more helpful to designers. It is not your enemy but your friend, your partner.
Designers who know how to use it will leverage it to achieve greater heights. ‘Space Travel’, image generated by Dream AI. For ordinary users, it is also very user-friendly; Dream AI can automatically arrange layouts, select colors, choose font sizes, and fonts, and match appropriate styles on its own. It must be said that Byte’s 2.1 image model indeed has something special; it is speculated that it has been pre-trained with some OCR and poster layout data, allowing the model to generate based on underlying logic. After the release of Dream 2.1, I have seen many people in the design community expressing pessimism, but I believe it is unnecessary. The notion that ‘design is doomed’ is not a new one; such opinions emerged two years ago when Midjourney first appeared, and Midjourney did not kill the design industry, so Dream may not either. Those who have been killed by MJ are probably no longer in the design industry by now. Two years have passed, and the use of MJ, SD, or liblib has become a basic skill in design. Now, we cannot live without AI. Just as everyone was inseparable from smartphones seven or eight years ago, smartphones and AI have become our foundational skills, subtly changing us, our work, our lives, and the efficiency of social organization. The empowerment brought by AI will grow stronger, regardless of anyone’s will. Times are changing, and what we need to do is change ourselves. AI is so interesting, let’s play with it, friends!
