I built dabuun (https://dabuun.com/), a tool that turns a single line of text into a ready-to-post social media video.
Instead of manually editing video timelines, dabuun takes a short description and automatically creates: - plot and script - scenes and images (anime or realistic) - voiceover - subtitles - final rendered video
It supports both Japanese and English and can format for YouTube, TikTok, and Instagram with the right aspect ratios. It also keeps character appearance consistent across scenes using AI-generated references, so your “main character” doesn’t change every cut.
This is meant for: - people who find video production exhausting or too time-consuming - people who only have a smartphone and no recording/editing gear - creators, educators, and storytellers who want to focus on ideas, not tools
Why I built it A while ago I made some YouTube videos manually, and it was mentally and physically draining. I wanted something that would let more people turn simple ideas into content without needing video skills or equipment. Long term, I’d like to make it easier for anyone, anywhere, to create monetizable content with just a phone and an idea—and maybe help reduce the current imbalance in the creator economy.
Tech stack - Frontend: Next.js - Infra: Vercel, GCP
Known limitations / next steps - Script quality still needs work; I want better story structure and pacing - I’d like to add “consulting” features that suggest how to improve your video for each platform - Longer term: evolve into a platform where creators can manage and publish content, not just generate single videos
I’d really appreciate any feedback, especially from people who actually try generating a video: - Does the output feel good enough to post as-is? - What feels off or low quality? - What would you need to trust this for your real content?
Thanks for reading, and I’m happy to answer any questions.