💬Generate SRT Captions
Turn spoken audio into subtitle files. Free, private, no signup — files never leave your browser.
First run downloads the AI model once, then it’s fast. 100% private — on-device. Fast = Whistle 16.9MB (7 languages) · Accurate = Whisper (99 languages).
Captions in one click.
Word-level timestamps from an on-device model, exported as a standard SRT file that every player reads.
How to use Generate SRT Captions
Captions are not only for accessibility — they are what lets a clip be watched on mute in a feed, and what most platforms reward. This page turns any recording into a ready-to-use SRT file: pick Fast mode for a 16.9 MB model that handles seven languages, or Accurate mode for Whisper-level coverage of 99 languages. The transcription runs on your own device, so an unreleased interview never touches a server.
- 1Drop in an audio or video file (MP3, WAV, M4A, MP4).
- 2Pick Fast (16.9 MB, 7 languages) or Accurate (99 languages).
- 3Transcribe, then copy the text or download the .srt file.
- 4Save the .srt next to your video with the same filename so players auto-load it.
Looking for the full version of this tool? Open Audio to Text — it covers every option, not just this one.
Use cases
What generate srt captions does
- ✓Word-level timestamps rendered into standard SRT cues
- ✓Fast mode: 16.9 MB, 7 languages; Accurate mode: 99 languages
- ✓Copy the transcript as plain text or download both formats
- ✓Fully on-device — no upload, no account
Specs
- Input
- MP3, WAV, M4A, MP4 (audio track)
- Output
- .srt subtitles and .txt transcript
- Engines
- Whistle 16.9 MB or Whisper (99 languages)
- Privacy
- On-device inference, zero uploads
- Cost
- Free, unlimited, no signup
Frequently asked questions
What makes an SRT file work with my video?
An SRT is a plain text file with a timecode, a separator line and a line of text per cue. Players load it automatically when the filename matches the video exactly — clip.mp4 needs clip.srt in the same folder. If captions do not appear, the filenames do not match or the file is not in the same directory as the video.
How long do the caption lines need to be?
For horizontal video, roughly 42 characters per line is the classic limit, and two lines per cue reads best. For vertical Reels and Shorts, keep lines under about 20 characters, because the safe area is narrower and the text is read at arm’s length. Our timestamps are word-level, so the captions track the speech rather than a rough paragraph split.
Should I use Fast or Accurate mode for captions?
Fast mode is enough for clean single-speaker recordings in English, German, French, Spanish, Italian, Dutch or Polish, and it loads a 16.9 MB model instead of a 145 MB one. Accurate mode is the better choice for accented speech, overlapping voices, background music, or any language outside those seven — it uses Whisper and covers 99 languages with auto-detection.
Can I get burned-in captions instead of a subtitle file?
This page gives you the subtitle file, which every player and every platform accepts. To burn captions into the picture, load the .srt in the full editor alongside your video, where you can style the text, position it and export a new file with the captions permanently rendered.
More free tools
Need more control? Open the full video + audio editor →Was this tool helpful?