Launched this week
HeadVox
Script → voiceover + word-aligned captions, open source
2 followers
Script → voiceover + word-aligned captions, open source
2 followers
Turn a text or Markdown script into a natural TTS voiceover plus SRT/WebVTT captions timed to the spoken words and a per-word JSON timeline. Optional one-step video mux with ffmpeg. Chinese, English, Japanese. Free, MIT, pip install.
Free
Launch Team
Render Sandboxes: a computer for every agentGive every agent its own machine. Starts in under a second
Promoted
Hi Product Hunt 👋
Making a short narrated video usually means three chores: generate the voiceover, hand-time the subtitles, then mux them onto the clip — and redo all three whenever one sentence changes. HeadVox makes that one command.
What it does
• Script (.txt or Markdown) in → voiceover.mp3 + .srt + .vtt + voiceover.json out
• Captions come from the TTS engine's real word boundaries, so they don't drift late in the video
• CJK-aware line breaking: Chinese/Japanese count as double width, so mixed text fits 9:16
• [pause 1.5s] anywhere for pacing; Markdown is cleaned into speakable text
• --video avatar.mp4 --music bed.mp3 gives you a finished MP4 (looped clip, ducked music, soft or burned subtitles)
• Per-word JSON for karaoke captions, lip-sync or B-roll cuts
• Free voices out of the box (no API key), offline test backend, custom engines in ~10 lines
What it isn't: it doesn't clone voices or render avatars — it's the focused layer between "my script" and "a publishable clip", and plays well with whatever editor you use.
Try it: pip install "headvox[edge]" — real samples on the Hugging Face page. Feedback on which languages and TTS backends to support next is very welcome!