How to add subtitles to a video from its audio

Captions are worth the effort. They make videos watchable with the sound off, they help viewers who are deaf or hard of hearing, and they tend to improve watch time and reach. The good news is that you can create accurate subtitle files from a video's audio for free, in your browser, without uploading anything. Here is the whole process.

How subtitles actually work

A subtitle file is just plain text with timestamps. Each entry says what to show and when. There are two common formats:

• SRT is the most widely supported format. Almost every video editor and player reads it, and most social platforms accept it for uploaded captions.
• VTT is the web format, used with the HTML5 video caption track. If you are publishing video on your own site, this is often what you want.

You do not have to choose in advance. The tool below produces both from the same transcription.

Step 1: get the audio out of the video

Subtitles are generated from speech, so you start with the audio. If your file is already audio, skip ahead. If you have a video file such as an mp4, extract or export its audio track as an MP3 or WAV. Most video editors have an export-audio option, and there are free tools that do just this. Because the audio comes from the same video, the timestamps will line up with the original.

Step 2: transcribe with timestamps

Open our free transcriber and drop in the audio. Keep timestamps on, which is the default, because the subtitle timing comes from them. If your video has more than one speaker and you want the captions to show who is talking, turn on speaker separation. Click Transcribe and let it run on your device. The first run downloads the model once and then caches it.

Step 3: check and fix the text

Open the Subtitles tab to see the timed lines. Skim for the usual suspects: misheard names, product terms, and numbers. Fix anything off in the in-page editor, since your edits carry into the export. Clean captions are worth a couple of minutes of review, because errors are much more noticeable on screen than in a plain transcript.

Step 4: export SRT or VTT

Download the SRT or the VTT file, whichever your tool needs. Then load it as a captions or subtitles track:

• In a video editor, import the subtitle file as a caption track and it will align with the video.
• On the web, reference the VTT file with a track element inside your video tag.
• On social platforms that accept caption uploads, upload the SRT alongside the video.

Tips for better captions

• Use the larger model option for tougher audio, such as strong accents or background noise. It is a bigger first download but more accurate.
• Keep lines readable. If a caption entry is long, your editor can usually split it. Short lines are easier to read on screen.
• For long videos, transcribe in segments. In-browser transcription is most reliable up to about forty minutes per file, so caption a long video in parts and combine the subtitle files in your editor.

Why do this in the browser

Plenty of sites will caption your video if you upload it. That is fine for public content, but for anything unreleased, internal, or sensitive, uploading is a real downside. Doing it in your browser keeps the audio on your device, costs nothing, and has no watermark or minutes cap. If you also need a readable transcript alongside the captions, you already have it: the same run gives you plain text and speaker labels. See our audio to SRT page to jump straight in.