How to make an SRT subtitle file
An SRT file is a small text file of numbered captions with start and end times. YouTube, Vimeo, Premiere, DaVinci Resolve and most players read it, and viewers can switch it on or off. This generator listens to your video or audio, writes the subtitles with word-level timing, and lets you fix them before you download. Nothing is uploaded: the speech model runs in your browser.
- 1
Add a video or audio file
MP4, MOV and WebM videos work, and so do sound files such as MP3, M4A and WAV, which is handy for podcasts and voice notes. Files can be up to 10 minutes long.
- 2
Let it transcribe
The first time, the tool asks before loading its speech model. It then finds the speech, transcribes it and times every word. Short clips are ready in seconds.
- 3
Check and correct
Read through the subtitles. Click a word to fix it, use Find & replace for names, and split or join lines so each one reads naturally. Choose how many words go on each subtitle.
- 4
Download SRT, VTT or TXT
Download an SRT file for most platforms and editors, a VTT file for web players, or a plain transcript. Your edits are in every format.
What’s inside an SRT file
SRT is the simplest subtitle format there is, which is why almost everything supports it. Each subtitle is a number, a time range and the text, followed by a blank line:
1
00:00:00,592 --> 00:00:02,672
Hey everyone, welcome back
2
00:00:02,672 --> 00:00:03,392
to the channel.Times are hours, minutes, seconds and milliseconds, with a comma before the milliseconds. VTT files look almost the same, with a WEBVTT line at the top and a dot instead of the comma. Our SRT vs VTT guide covers which one each platform prefers.
Where to use your subtitle file
- YouTube: in YouTube Studio, open your video’s Subtitles section and upload the file. Uploaded subtitles replace YouTube’s automatic ones, so names and terms are spelled the way you want.
- Vimeo, Wistia and course platforms: most accept SRT or VTT in the video’s settings.
- Video editors: Premiere Pro, DaVinci Resolve and Final Cut Pro can import SRT as a caption track that you can restyle.
- Podcasts and transcripts: the TXT download is a clean transcript for show notes, blog posts or quotes.
For TikTok, Reels and Shorts, captions are usually burned into the picture instead. Use Auto Captions for that: it is the same tool with styles and video export.
How good is the timing?
Each word gets its own start and end time. The tool first finds exactly where speech starts and stops in your audio, transcribes the speech in pieces that break at natural pauses, then trims each word so subtitles don’t linger into silence. In our tests on recorded speech, subtitle ends landed within about a seventh of a second of the voice on average.
If one line still feels early or late, the + and − buttons move it by a tenth of a second without overlapping its neighbors.
Private by design
Transcription services usually upload your recording and keep it on their servers. Here, the speech model is loaded into your browser once and does the work on your device. Unreleased videos, interviews and client recordings never leave your computer. You can remove the saved model any time from the Saved AI models page.
Frequently asked questions
Is the SRT generator free?
Yes. SRT, VTT and TXT downloads are free, with no sign-up, no watermark and no limit on how many files you make.
Can I make subtitles from an audio file?
Yes. MP3, M4A, WAV and other common sound files work, which makes it useful for podcasts, interviews and voice notes.
Which languages does it support?
English only for now. The speech model is built for English, and we would rather not offer other languages until they work well.
Should I download SRT or VTT?
SRT works almost everywhere, including YouTube, Vimeo and video editors. Choose VTT for web video players and sites that ask for it. Both contain the same subtitles.
Are the subtitles accurate?
On clear speech they are very accurate, but no automatic transcription is perfect. Names, slang, accents and background music cause most mistakes, so read through and fix anything before downloading.
Is my recording uploaded anywhere?
No. The speech model runs in your browser and your file never leaves your device. Only the model itself is loaded, once, from our site.