Frameloom

Generate subtitles from speech automatically

Typing subtitles by hand takes longer than the video itself. Frameloom listens to your video, writes the subtitles with the timing of every word, and lets you fix anything it misheard before they go on the picture. The speech recognition runs on your own device, so the recording never leaves it.

Free to try: a 1-minute 1080p video costs 7 tokens, and you get 30 every week.

Recognised on the device, laid out into readable subtitles, burned in · real result on a sample clip

How to do it

  1. Open Auto subtitles and drop your file in, or pick it from your device. It stays on your device.
  2. Check the settings. Choose the spoken language, Sentences or Short bursts for Reels and TikTok, and whether you want the video with subtitles or just the .srt file.
  3. Press Run, then preview and download the result.

Subtitles that read well

Recognised words are grouped into subtitles the way professional subtitlers do it: a new subtitle at each sentence and pause, lines balanced and kept short enough to read, commas used as natural break points, and each subtitle on screen long enough to follow.

Before anything is burned in you get the full list of subtitles next to a live preview. Correct a name, change a word, remove a line, then make the video. You also get the .srt file, for YouTube or any editor.

The first time, your browser downloads the speech model (80 MB, or 250 MB for Accurate). After that it's cached. On computers with a modern graphics card it runs on the GPU and takes seconds.

PrivateYour file is processed in your browser and never uploaded.
No installWorks in Chrome, Edge, Firefox and Safari on any computer.
Smart settingsSettings are chosen for your specific file, not one-size-fits-all.
Free weekly30 tokens every week with a Google account. No card needed.

Questions

Which languages work?
English, Romanian, Russian, Ukrainian, Spanish, French, German, Italian, Portuguese, Polish, Dutch and Turkish. For languages other than English, choose Accurate.
Is my audio uploaded?
No. The recognition model runs in your browser. Only the recognised words and their timings go to our server, which lays them out as subtitles and doesn't keep them.
Can I just get the subtitle file?
Yes. Choose Subtitle file only and you get a standard .srt that works in YouTube, Premiere, DaVinci Resolve, CapCut and media players.
What if it gets a word wrong?
You check every subtitle before the video is made and can edit or delete any of them.

Related