rockbenben/img-prompt

What it solves

IMGPrompt removes the language barrier for non-native English speakers creating AI art. It allows users to build complex, high-quality English prompts for image and video generation models without needing to be fluent in English, by providing a curated, multilingual library of keywords.

How it works

Users navigate a three-step workflow: first selecting an object (like a character or scene), then an attribute (such as lighting or camera angle), and finally clicking specific tags. The app translates these native-language selections into precise English terms that AI models understand. It also includes a real-time translation utility for free-form text and a prompt output panel for final refinement.

Who it’s for

Non-native English speakers who use AI image and video generators like Midjourney, FLUX, Kling, and Stable Diffusion and want more control over their artistic output without wrestling with English vocabulary.

Highlights

• 5,000+ Curated Tags: A massive library of styles, artists, and compositions organized by category and translated into 18 languages.
• Visual Discovery: Hovering over tags shows a preview image, with a full lightbox for zooming and rotating.
• Multi-Model Compatibility: Works with a wide range of tools including GPT-Image-2, Nano Banana, Midjourney, FLUX, Seedance, Veo, and Kling.
• Smart Output Panel: Features inline template buttons for common quality boosters, auto-translation for typed text, and deduplication of tokens.
• Cross-Platform: Available as a web app or as native clients for Windows, macOS, and Linux.