Ai Art Generation has been evolving at a wild rhythm, and Google only threw another great contender to the mixture through its Gemini Flash 2.0. You can play with the new image creation tool in the Google artificial intelligence study.
Gemini Flash is, as the name implies, very fast, especially faster than Dall-E 3 and other image creators. That speed can mean lower quality images, but that is not the case here, especially because all changes and updates to the image production capacity of the model. Even so, if you really want good results, you must know how to talk to AI. After a lot of proof and error, I have gathered five tips to obtain the best absolute art of Gemini Flash 2.0. Some of these may seem similar to the advice on other AI art creators, because they are, but that does not make them less useful in this context.
Tell a story
The new most interesting feature for the creation of images of Gemini Flash is that it is not only good for unique illustrations, it can actually help you create a visual story generating a series of images related to style, configurations and consistent moods.
To start, you just have to ask you to tell you a story and how often you want an illustration to go with the action. The result will include those images that accompany the text.
For my project, I asked the AI to “generate a story of a heroic dragon that protected a fairy queen of an evil magician in a 3D cartoon animation style. For each scene, generate an image.” I saw the previous beginning to appear. And, if there is a problem, you can rewrite any of the bits of history and the model will regenerate the image accordingly.
Be super specific
If you tell Gemini to make “a dog in a park”, you may get a blurred Golden Retriever sitting in a vaguely green place. But if you say: “A spongy Retriever sitting on a wooden bench in Central Park during the fall, with red leaves and oranges scattered on the ground,” you get exactly what you are imagining.
The models of ia prosper in details. The more I provide, the better your image will be. Then, for the image above, instead of asking for a futuristic appearance city, I requested “a retro-fouturist urban landscape at sunset, with neon signs that shine in pink and blue, flying cars in the sky and people who walk with retro-fouturous style attire.” Seven seconds later, the result came.
Get conversation
One of my favorite things about the new Gemini Flash is that you can talk with him without losing much of the speed. That means you don't have to do everything well at once. After generating an image, you can literally chat with AI to make editions. Do you want to change the colors? Add a character? Make lighting more humor? Just ask.
In the image above, I started asking “a cozy reading corner with a fireplace, shelves full of novels and a great comfortable armchair.” Then I refined it by asking him to “do it at night with soft and warm lighting”, then I kept asking him to “add a sleeping cat in the armchair” and ended up requesting the AI ”give the room a vintage and Victorian aesthetic.” The final result on the left is almost exactly like what I imagined, and makes Gemini feel an art assistant, one capable of adapting to what I want without starting from scratch every time.
Gemini Flash coincides chatgpt
Google has boasted that Gemini is full of knowledge of the real world, which means that you can obtain historical precision, realistic cultural details and realistic images if you ask. Of course, that requires being specific. For example, if you request it for “a Viking warrior”, you may get something that looks more like a game of Thrones character. But if he says: “A historically precise Viking warrior of the ninth century, with detailed chain cane armor, a round wooden shield and a traditional Nordic helmet,” will get something much more precise.
As proof, I asked the AI to make “an old Mayan city at dawn, with imposing stone pyramids, lush jungle environments and people dressed in traditional Mayan garments.” It is not perfect, but it looks much more like the real thing than the previous versions, which sometimes returned with an almost Egyptian pyramid.
Write fast
Most IA image models have long fought with the representation of the text, turning the words into illegible scribbles. Even the best models of today who can do it, take a bit to do it and do it well can take some attempts. But, Gemini Flash is surprisingly good to integrate the text into images quickly and readable. However, being very specific can help.
This is how I generated the image above asking the AI to “make a vintage -style travel poster that says 'Visit London' in black and retro typography, with a stylized illustration of the city.”
