Storybook Maker: A Generative AI Based Approach for Automated Storybook Generation

Abstract

Generative Artificial Intelligence (Generative AI) has rapidly changed digital content creation by allowing automated generation of text, images, and multimedia content from natural language prompts. Nevertheless, current storytelling platforms that are AI-based are typically limited to a single function, thus requiring the user to use multiple apps for illustration, translation to multiple languages, narration, and layout. In this contribution, we introduce Storybook Maker, a web-based generative AI architecture for the automated generation of digital storybooks that integrates several AI services into a common platform. The proposed tool employs Gemini 3 Flash Preview for story generation and FLUX.1-dev for text-to-image synthesis while allowing translation to any other language, text-to-sound conversion, and PDF storybook generation. Software implementation will be conducted using Python, Django, and MySQL with modular architecture. An extensive evaluation was performed using various LLMs and diffusion-based text to image models. The preliminary findings show that Gemini 3 Flash Preview is the best-performing system in terms of story generation, according to BLEU, ROUGE, BERTScore, and human scores. Conversely, FLUX.1-dev is better than any other known diffusion-based model if considered in terms of CLIP Score, ImageReward, SSIM, PSNR, and Human Evaluation. Thus, an efficient, complete website was created for AI supported digital storytelling that allows making story creation easier.