I built a tool to extract hardcoded subtitles from videos, running entirely in the browser

Hey everyone, I wanted to share a tool I’ve been working on recently! It’s a web tool that extracts hardcoded (burned-in) subtitles straight from your videos and turns them into clean, editable text and .srt files. Whether it's a downloaded film, a drama episode, anime, or a short clip, it reads the text directly off the video frames so you never have to pause and type out dialogue by hand. It works across a wide range of languages including English, Chinese, Japanese, Korean, and various Latin-script languages. The tool is completely privacy-friendly and runs 100% locally in your browser. Your videos are processed entirely on your machine and are never uploaded to any server. No account or sign-up is required. It also automatically indexes every extracted line into a searchable local dialogue library. You can type in any word or phrase to search across all your processed videos and jump straight to the exact scene and timestamp where it was spoken. Tech stack under the hood: Tanstack Start PP-OCRv6 Sqlite wasm The app is called SubKu, and it’s completely free to use: subku.app/en