
區塊先生 🐡 ⚠️ (rock #58)|Mar 26, 2025 13:23
Ghibli Festival ChatGPT
OpenAI recently updated the GPT-4 Turbo (with Vision), which now supports text to image style description generation. This image is implemented through the DALL · E 3 API and combined with prompt engineering to simulate Studio Ghibli style.
The main method behind generation is through the Dalle-3 model, combined with ChatGPT's image input module. There is no need to write code yourself, but if you need to manually implement it, you can refer to OpenAI's openai python repo and use it with OpenAI Image.create and openai ChatCompletion.create makes combinations.
This feature is currently only available on ChatGPT Pro (GPT-4) and supports image uploading and image conversion generation.
Prompt: Please convert this photo into a Ghibli style animated illustration. Retain the original composition and character characteristics, and present them with soft brushstrokes and warm tones. The background is delicate and realistic, creating images with magical realism and nostalgic atmosphere like "The Hidden Maiden" or "Howl's Moving Castle".
Share To
Timeline
HotFlash
APP
X
Telegram
CopyLink