Auto Subtitles

Turn speech into captions in its own language, edit them and add them to the video.

Processed in your browser — files never leave your device

Drop a file here

or

Select file

About Auto Subtitles

Turn the speech in a video into captions written in the language that is spoken — Hindi in Devanagari, Tamil in Tamil script, English and many more. Correct the words, pick a style, then add the captions to the video or download an SRT file. Speech recognition runs in your browser.

How Auto Subtitles works

Auto Subtitles listens to the speech in your video with Whisper, OpenAI's open-source speech-recognition model, running in your browser. It detects the spoken language (or uses the one you choose) and writes the captions in that language's own script, with the time each line is spoken.

You can correct every line before using it. Then the captions are drawn into the video with a font made for the script — Noto Sans for Latin, Devanagari, Bengali, Gurmukhi, Gujarati, Tamil, Telugu, Kannada and Malayalam, and Noto Naskh Arabic for Urdu and Arabic — or you can download them as an SRT file for YouTube or a video editor.

The speech model is downloaded once from the Hugging Face CDN (about 250 MB for Accurate, 77 MB for Fast) and kept by your browser; your video itself is never uploaded. Recognition works best on a computer: phones may be very slow or run out of memory.

When to use Auto Subtitles

  • Add captions to reels and shorts — most people watch them without sound.
  • Make lectures, interviews and talks easier to follow, in Hindi, English and other languages.
  • Create an SRT file to upload as subtitles on YouTube.
  • Make your videos accessible to people who are deaf or hard of hearing.

How to use Auto Subtitles

  1. Add your video. Drop the video into the box above.
  2. Generate and correct. Choose the language (or leave it on Detect) and the accuracy, click Generate captions, then fix any words in the list.
  3. Add to the video. Pick the caption style and click Add captions to video — or click Download .srt.

Options explained

Spoken language
Detect automatically, or choose it if detection picks the wrong one — for example for very short clips.
Accurate
The larger model: much better for Hindi and other Indian languages. A 250 MB download the first time.
Fast
The smaller model, good for clear English. A 77 MB download the first time.
Style
White or yellow text with a black outline, or white text on a dark box; small, medium or large; at the bottom or the top.

Frequently asked questions

Which languages are supported?

Speech recognition understands about 100 languages. Captions can be drawn into the video for languages written in Latin, Cyrillic, Greek, Devanagari, Bengali, Gurmukhi, Gujarati, Tamil, Telugu, Kannada, Malayalam and Arabic scripts; for others you can download the SRT file.

How accurate are the captions?

Clear speech without loud music gives good results, especially with Accurate. Always read them through: names, mixed Hindi-English and noisy parts can need corrections.

Is my video uploaded?

No. The speech is recognised in your browser. Only the speech model itself is downloaded, once, from the Hugging Face CDN.

Is Auto Subtitles free?

Yes. Auto Subtitles on DocEpic is free, with no sign-up, no watermark and no limit on how often you use it. The site is supported by ads.

Are my files safe with Auto Subtitles?

Yes. Auto Subtitles runs entirely in your web browser, so your files never leave your device and are never uploaded to us.

Do I need to install any software?

No. Auto Subtitles works in any modern browser on Windows, macOS, Linux, Android and iPhone.

Is there a file size limit?

Files up to 1024 MB are accepted. Because the work happens on your device, very large files depend on your device's memory.

Related tools