fareskirouane5-lang/Open-Generative-AI — reverse-engineered prompt

Build me an open source AI creative studio that people can run in the browser or as a desktop app. I want it to feel like a simple alternative to paid AI video tools, with separate areas for Image, Video, Lip Sync, and Cinema workflows.

Users should be able to type a prompt, upload reference images or audio, pick from lots of models like Flux, Midjourney, Kling, Sora, Veo, Wan, and lip sync models, then generate and download the results. Please include a clean modern interface, generation history, model settings, progress states, error messages, and a settings page for API keys plus local model options where possible.

Make it self hosted, easy to customize, and don’t add extra prompt blocking beyond what the model provider requires. The desktop version should feel polished and installable. Please wire up realistic demo data or placeholders where real API access is needed, and look up current docs online if you need to.