AZTools.us
English

Auto Subtitle Generator & Audio Transcription (SRT, VTT)

Video preview with the subtitles SRT / VTT with timestamps · 19 languages

Public-domain LibriVox audiobook recording, downloaded to your browser only when you click.

Subtitle Generator settings

Load a file to unlock this tool’s controls:

Video preview with the subtitles

Runs directly on the web inside your browser. Your original file is never modified or uploaded.

Subtitles make videos easier to follow and searchable, and a written transcript is handy for notes, quotes and translations. This tool listens to a video or audio file with Whisper, an open speech-recognition model, running in your browser, and gives you subtitles with timestamps as .srt, .vtt or plain text. Your file is not uploaded.

Updated

How to generate subtitles

  1. Drop a video (.mp4, .mov, .webm, .m4v) or an audio file (.mp3, .wav, .m4a, .aac, .ogg, .opus, .flac) onto the workspace. The tool reads its sound; files up to one hour long are accepted.
  2. Choose the “Language”: “Detect automatically”, or the language that is spoken, which is more reliable for short clips and accents. Under “Output”, keep “Same language” or choose “Translate to English”.
  3. Press “Start”. The first time, the speech model (about 106 MB) is downloaded from aztools.us and kept in your browser; later runs start immediately. The progress line shows how much of the recording has been transcribed. “Stop” cancels.
  4. Read the result in the preview. For a video, play it: the subtitles are shown on the picture so you can check the timing.
  5. Under “Format”, choose SRT, VTT or TXT, then export. Switching the format does not transcribe again.

Which format to choose

Export formats
FormatUse it for
.srtYouTube, Facebook, TikTok uploads and most video editors (Premiere, DaVinci Resolve, CapCut)
.vttWeb video players (the HTML <track> element) and some learning platforms
.txtA readable transcript without timestamps; a new paragraph starts after a pause of two seconds

Accuracy and speed

  • The model is Whisper base, an 8-bit version that runs on the processor. Clear speech in English and other major languages comes out well; names, rare words and heavy accents need checking.
  • Lines are kept to about seven seconds and 84 characters, the usual limits for subtitles.
  • Background music, several people talking at once and long silences can produce missing or invented lines. The tool drops tags such as [Music] and the repeated loops the model produces on music, and says so when a file seems to be mostly music.
  • On a recent computer, transcribing takes roughly a quarter to half of the recording’s length. Phones and older computers are slower; keep the tab open while it works.

Subtitle Generator: common questions

Which languages are supported?
The list has English, Vietnamese, Chinese, Japanese, Korean, Spanish, Portuguese, German, French, Italian, Russian, Indonesian, Thai, Hindi, Arabic, Turkish, Dutch, Polish and Ukrainian. Automatic detection can recognise other languages Whisper knows as well.
Can it burn the subtitles into the video?
No, it creates a separate subtitle file. Upload the .srt together with the video, or add it in your video editor, where you can also style and position the text.
Is my video uploaded?
No. The sound is decoded and transcribed in your browser tab. Only the speech model and its runtime are downloaded, once, from aztools.us.