Submagic charges $18–39 a month to run Whisper and ffmpeg on your clips. Captionly is $34 once. Drop a video, run Whisper locally for word-level timestamps, pick a caption style, fix any misheard words inline, and burn a ready-to-post MP4 — 100% on your machine. Same pipeline, your hardware, zero subscription.

• Local Whisper transcription — whisper.cpp with word-level timestamps; the engine and model download once on first run, then run fully offline.

• Karaoke word-highlight captions — Real ASS timing tags per word for the TikTok and Reels look, rendered by ffmpeg exactly as previewed.

• 3 style presets — Karaoke Highlight, Bold Center and Classic Subtitle, plus font size, position and highlight color controls.

• Inline editing with re-sync — Click a caption, fix the words, and timings redistribute proportionally across the cue with no re-transcribing.

• Keyword to emoji decoration — Optional mapping turns words like money into an emoji, with 50+ built-in mappings.

• Burn-in export — ffmpeg ass filter to H.264 MP4 with audio stream-copied and live progress.

• SRT and ASS export — Export subtitle files for YouTube uploads or further editing.

• 100% local — Your footage never leaves your machine — no accounts, no telemetry.

Runs as a Windows desktop app.

What's included:

• Full source code (MIT): https://github.com/bensblueprints/video-clip-captioner

• Lifetime v1 updates

• Setup guide (README quick start)