How Dynamic Caption Styling Works
AnyTool adds animated captions to a video entirely in the browser. It transcribes speech with a self-hosted Whisper model (WebGPU with WASM fallback) to get word-level timestamps — or accepts a pasted SRT/VTT for an instant path — then renders captions per frame on a canvas with word-by-word reveal and an active-word karaoke highlight, scaled with a pop animation. Frames are encoded to H.264 MP4 with WebCodecs (hardware-accelerated), muxing the original AAC audio; a self-hosted libass/ffmpeg path burns clean static styles as a fallback. No server, no CDN.
- Runs client-side — video, transcript and model stay on device
- Self-hosted Whisper transcription (or paste an SRT/VTT)
- Animated word-by-word + karaoke active-word highlight
- libass/ffmpeg H.264 burn-in (self-hosted single-thread core)
- Captions burned into a real MP4, audio preserved
- No watermark, no sign-up, no length or count caps
