ChatGPT's new native image generation capabilities have been causing a stir on OpenAI's servers as users eagerly request the chatbot to convert their real-life photos into a Studio Ghibli-style art. While these image-generation features provide more detailed and contextual visuals compared to other chatbots, they remain limited due to high demand. Free users, for example, can only generate three images, while even paid users face usage limits.
In contrast, xAI’s Grok chatbot may not generate the most accurate images, but it offers a more extended image upload and creation limit (specific limits haven't been provided by the company). But what if you could enhance the image creation process by using ChatGPT to help Grok produce more nuanced visuals? Here's a step-by-step guide.
How ChatGPT Can Assist in Creating Better Images with Grok
Though modern large language models (LLMs) like Grok can generate images based on natural conversational prompts, users often face disappointments due to missing details or the chatbot hallucinating key elements of the image. To overcome this, creating highly detailed prompts that account for aspects like context, subject, background, theme, color palette, atmosphere, and art style is essential.
Avoiding ambiguity is critical in ensuring the chatbot generates a more accurate and coherent image. This is where tools like ChatGPT or even Gemini come in handy to refine the prompt, considering user preferences and minimizing ambiguity.
We tested using Grok to generate a Studio Ghibli-style image of three famous Indian cricket captains. While the initial result was unsatisfactory, the outcome drastically improved when ChatGPT was used to generate a more specific text prompt. The first image, created directly by Grok, featured wrong jersey patterns, inaccurate faces, and a weak background. However, when ChatGPT was asked to refine the prompt, the second image generated by Grok showed improved accuracy in facial resemblance, Ghibli-style effects, and even correct jersey patterns. Despite some errors in the franchise logos, the team logos were accurate.
How to Generate a Studio Ghibli-Style Image Using Grok with ChatGPT’s Help:
Open the ChatGPT website or app and describe the image you want to generate, providing as much detail as possible.
Ask ChatGPT to craft a text prompt for Grok to generate the image.
Open the Grok app and input the text prompt generated by ChatGPT.
Your desired image will be ready in seconds. If you need adjustments, you can ask Grok to make further edits using ChatGPT’s help.
Grok operates on xAI’s latest foundation model, Grok 3, which was launched last month. Initially available only to X subscribers, it was later opened to free users due to intense competition from Chinese AI firms like DeepSeek and Qwen. Despite Grok 3’s impressive photorealistic and detailed image generation capabilities, it was soon overshadowed by the more nuanced image generation features from Google and, more recently, ChatGPT.
What is Studio Ghibli?
Studio Ghibli is a renowned Japanese animation studio founded in 1985 by Miyazaki Hayao, Takahata Isao, and Suzuki Toshio. Known for its exceptional hand-drawn animation and captivating storytelling, Studio Ghibli has produced iconic films such as My Neighbor Totoro, Spirited Away, Howl's Moving Castle, Kiki's Delivery Service, and Princess Mononoke. The studio's work is celebrated for its dreamlike landscapes, soft color palettes, and deeply human narratives. Ghibli’s hand-drawn animation techniques have long been considered the gold standard in traditional animation.
