Hi Product Hunt 👋 We built this browser workspace around a practical question: when an AI video comes out “almost right,” how should a creator know what to change next? The problem might be the prompt, a reference image, a motion video, audio guidance, aspect ratio, duration, resolution, or the interaction between them. When these inputs are scattered across different steps, iteration quickly becomes difficult to understand. The current workflow brings text, image, video, and audio references together with model-specific output controls, credit estimates, and generation history. Creators can plan a generation, review the result, and make the next change with more context. Our goal is not to promise a perfect finished video in one click. It is to make AI video creation clearer, more reviewable, and easier to iterate. We’d especially value feedback on three questions: 1. Which input do you usually start with? 2. What most often makes you regenerate? 3. Would parameter differences, side-by-side playback, or failure notes make iteration more useful? Thanks for checking it out. Specific feedback on anything unclear or frustrating is very welcome.

Wan 3.0

Plan, generate, and iterate AI videos in one workspace

Plan, generate, and iterate AI videos in one workspace