Table of Contents 64q43
Amidst the bombast trends of the anime style photos, you might have wondered: how does the Chat GPT makes images in the style of Studio Ghibli? He couldn't before? Well... it turns out that the new generative imaging capability of OpenAI is really new and has everything to do with a major change in the way the company's artificial intelligence model works. Understand, in this text, how the Chat GPT managed to replicate the art of Studio Ghibli and the main differences compared to other models. 4q606q
How does ChatGPT create Ghibli-style images? k165f
To understand how the Chat GPT began to be able to generate images like those created in trends do Studio Ghibli, it is necessary to know the main changes brought about by OpenAI in the model GPT-4o, read below.
What has changed in the GPT-4o model? 6r5943
The generation of images in the GPT-4o is based on architecture Transform, which analyzes patterns and associates textual descriptions with visual elements. Unlike “diffusion models,” which start with a random image and progressively refine it, GPT-4o adopts an autoregressive method, building the image in sequential steps, which improves visual coherence and fidelity to the desired style.
As the accuracy and rendering of text within images improves, and the ability to understand and follow detailed prompts improves, Chat GPT has increased its capacity to manage the number of distinct objects in a single scene (up to 20, compared to previously between 5 and 8 objects) and can maintain the correct relationships between them.
Compared to previous models such as the DALL-E3, which used pipelines (set of steps or processes) separate for different types of media, the GPT-4o takes a unified approach, enabling faster responses and more consistent visuals with textual commands, eliminating the need for multiple intermediate processes. This means the model can better interpret the nuances of -provided prompts, ensuring that the visual output is more aligned with the original intent.
All this power comes with a controversial problem: with a greater demand for graphic processing due to the explosion of fashion on the internet, OpenAI needed to limit the usability of the tool for a few days, especially for free s of Chat GPT, because their GPUs were “melting,” according to the company’s CEO, Sam altman. Despite the inconvenience, the generation of advanced images has already been released by the company for s of the free plan, as announced by the executive.
How does the update enable Ghibli-style images? 5s43e
But what does this all mean for AI's ability to capture the aesthetics of Studio Ghibli? It turns out that with the new update, the GPT-4o It begins to better understand style patterns, from massive examples of images, and learns to replicate these characteristics based on commands provided by the .
Although the OpenAI did not explicitly state whether their training dataset included specific works of art Studio Ghiblias the A Viagem de Chihiro ou The Boy and the Crane – straddling a gap that differentiates the reproduction of studio styles from those of living artists – the model's proficiency in replicating the studio aesthetic strongly suggests that he has been exposed to a large number of examples of the films and related styles.
Through this exhibition, the GPT-4o learn the main visual characteristics that define the style of Studio Ghibli, including: the feel of hand-drawn animation, often reminiscent of traditional paper-based animation; backgrounds that evoke a sense of wonder and tranquility; expressive characters typically featuring large, soulful eyes; a soft, natural color palette dominated by pastel and muted tones; detailed environments that often blend fantasy elements with realism
Furthermore, the GPT-4o can also process images as input, allowing s to transform existing visuals or use them as a basis for generating new content. This way, in addition to not having to say every detail about the artistic style they want to reproduce, s also don't have to describe themselves or the photo they want to recreate, simply inserting a photo into the chat and asking the AI to reproduce it with the animation style – as you can see in our article on How to Make Ghibli Style Pictures with ChatGPT, with hints and prompts.
Can other AIs reproduce Studio Ghibli's style? 5h595i
Create new images using chatbots is not a recent novelty, however, the capacity achieved by updating the Chat GPT in analyzing and reproducing the details suggested by the prompt, they replace the AI OpenAI ahead of its main competitors, such as Gemini, from Google, Grok, XAi e Copilot, Microsoft, in of image generation.
See the difference between the same image transformed into a drawing with a line inspired by Studio Ghibli following the same prompt:
Crie uma imagem, inspirada nessa foto, com base no estilo do Studio Ghibli
In addition to better capturing the details of the original image, one of the greatest advances in generative imaging GPT-4o This is evident in the ability to reproduce the logo and letters on the shirt, which is very difficult in other models.
Enquanto or Grok manages, in a way, to recreate characteristics of the style of Studio Ghibli, even though it suffers from the detail of the image, the Gemini goes completely against the grain, resulting in generic art. It is worth ing that the Google already invests in an update of the generative images of Gemini, enabling the creation of complex images and photo editing – tools that can be tested through AI Studio. O Copilot, Microsoft, failed to respond to the prompt.
Have you already tested the tool to enter the Studio Ghibli trend? Tell us in the comments.
See also other features
Text revised by Alexandre Marques in 01 / 04 / 2025.
Sources: The Times of India, Hindustan Times, Economic Times
