A web interface for open speech models. Upload a video, get subtitles. No GPU needed. No setup. Just works.
Drop your file here
Video or audio • MP4, MOV, MP3, WAV, etc.
Input → Model → Output. That's the whole pipeline.
Upload a file or paste a link. We extract the audio track.
MP4, MOV, MP3, WAV, YouTube, Vimeo
Whisper or AquaVoice processes the audio server-side.
Runs on our GPUs, not yours
Text is synced to timestamps. Edit if needed.
Millisecond precision
Download subtitles in your preferred format.
SRT, VTT, TXT, JSON, ASS
What the captions look like after processing.
Pretty simple. Gets the job done.
So I've been using this tool for my podcast.
You just drop the audio file and it runs Whisper.
The timestamps are already synced when you download.
Pretty simple. Gets the job done.
Convenience. That's the whole pitch.
Whisper runs on our servers. Your laptop doesn't need to be a beast.
No Python, no pip install, no CUDA drivers. Just open the page.
A 10-minute video takes about 2 minutes. Not instant, but not slow.
Whisper supports 50+ languages. Detection is automatic.
Standard caption tool features. Nothing fancy, everything useful.
Open models, proven accuracy. No proprietary black box.
Auto-detected. English, Spanish, Japanese, Arabic, you name it.
The models handle sentence breaks and punctuation on their own.
Fix names, adjust timing, tweak wording before export.
SRT, VTT, TXT, JSON, ASS. Pick what your tool needs.
Aligned to the audio with millisecond precision.
Works reasonably well even with background noise.
Queue up multiple files. Walk away.
If it has audio, we can transcribe it.
Transparent about the tech. Honest about what we do with your data.
Whisper is open-source from OpenAI. AquaVoice is open-weight. No mystery APIs.
We process your video, generate captions, and wipe the file. We don't keep copies.
Use it without signing up. Pay only if you want higher limits or priority.
An indie tool, not a VC-funded platform. We use it ourselves.