Repères Captions

Guide · Subtitle files

Generate an .srt file from a video, for free

An .srt file holds your subtitle text and its timings. You then load it into YouTube, a video editor or a video player. Repères generates it automatically from speech, in your browser, without uploading your file.

Open the free tool

The tool runs in your browser; its interface is in English.

In short

Drop in the video (or an audio file), click "Transcribe", review, then "Download .srt". The .srt export is free, unlimited and has no watermark. A .ass export with a karaoke effect is also available.

Step by step

  1. Open the tool on a computer, ideally in an up-to-date Chrome or Edge: they support WebGPU, which speeds up transcription.
  2. Drop in the video or the audio. The tool also takes audio files (MP3, WAV, M4A, OGG), handy for a podcast. The file stays on your device.
  3. Pick the spoken language and the model. 8 languages are available (English, French, Spanish, German, Italian, Portuguese, Dutch, Arabic). Model "Accurate" (whisper-small, ~250 MB) or "Fast" (whisper-base, ~90 MB), downloaded once and then cached.
  4. Click "Transcribe". The Whisper speech recognition model detects every word and the moment it is spoken.
  5. Review. Every line is editable; when you fix a line, its start and end times are kept.
  6. Choose the segmentation with "Words per line" (1 to 5): it sets the length of each block in the file. Untick "Uppercase" if you want normal casing.
  7. Click "Download .srt" (or "Download .ass (karaoke)"). The file is named after your video, followed by "-subtitles".

What the file looks like

Each block has a number, a start and end time (hours:minutes:seconds,milliseconds), then the text:

1
00:00:00,400 --> 00:00:01,406
Hi everyone,

2
00:00:01,406 --> 00:00:02,900
here is the result.

Timings come from Whisper's word-level timestamps: a block starts when its first word is spoken and ends shortly after the last one, without overlapping the next block.

And the .ass karaoke file?

The .ass file keeps your colors, size and position, in 1080×1920, with a karaoke effect (\k tags): each word changes color when it is spoken. It plays in software that supports the ASS format (VLC, Aegisub, or ffmpeg to burn the subtitles in). It expects the Montserrat Black font; if it isn't installed, a fallback font is used.

Where to use your .srt

  • YouTube: in YouTube Studio, under Subtitles, upload the file for the video's language.
  • Video editors: DaVinci Resolve and Premiere Pro import .srt files onto a subtitle track.
  • Video players: VLC loads an .srt placed next to the video with the same name.

Limits, honestly

  • A recent desktop browser is required. Tested on desktop Chrome and Edge (WebGPU). Firefox and Safari have not been fully tested yet; without WebGPU the tool falls back to WebAssembly, which is slower. On phones it may fail for lack of memory.
  • First run. The model is a one-time download of ~90 MB (Fast) or ~250 MB (Accurate), then it is cached.
  • Processing time depends on your hardware. WebGPU uses your graphics card; the WebAssembly fallback runs on the CPU and is noticeably slower. Video export runs in real time (a 60 s video takes about 60 s).
  • The AI sometimes gets it wrong (names, numbers, acronyms). Review and fix the text before exporting.
  • Output format. The video export is vertical 1080×1920: MP4 on recent Chrome/Edge and Safari, WebM on Firefox.
  • Short videos. Built for short-form content (up to 3 to 5 minutes). Beyond that, the browser may run out of memory.
  • Watermark on the free video export. Exported videos carry a small "Captioned with Repères" at the bottom right; the Pro license (€19, paid once) removes it. The .srt and .ass files never have a watermark.

FAQ

Is the .srt free and without watermark?

Yes. The free version's watermark only applies to the exported video; .srt and .ass files never have one, and there is no usage limit.

Is the .srt timed word by word?

No: an .srt holds blocks of a few words. Word-level timing drives the video animation and the .ass karaoke file.

Can I make an SRT from an audio file?

Yes: MP3, WAV, M4A and OGG files are accepted, as well as videos.

Can I edit the file afterwards?

Yes, an .srt is a plain UTF-8 text file that opens in any editor. The easiest is still to fix the text in the tool before exporting, since timings are kept.

Which languages?

English, French, Spanish, German, Italian, Portuguese, Dutch and Arabic. Pick the spoken language before transcribing; there is no translation.

Try it now

Free, no sign-up, and your video never leaves your computer.

Open the tool

Other guides