Auto Subtitles for Your Videos

Generate subtitles automatically, free and with no watermark

More than eight in ten videos on social media are watched with the sound off: without captions, your message goes unnoticed. Yet typing subtitles by hand takes hours. This auto subtitle generator does the work for you: the AI transcribes the speech, then the text is split into lines synced with the voice. Unlike many online captioning sites, there is no watermark on the video, no sign-up and no paid minutes cap. It works with MP4, MOV, WebM or MKV videos, and even with a plain audio file (MP3, WAV, M4A) if all you want is the subtitle file. It understands English, Spanish, French, Arabic, German, Italian, Portuguese, Japanese, Chinese and other languages.

Auto captions for TikTok, Reels and Shorts, burned into the video

TikTok, Instagram Reels, YouTube Shorts and WhatsApp don't accept a separate subtitle file: the text has to be “burned” into the picture. Choose the TikTok style (big white uppercase text with a black outline, in the center of the screen) or a calmer style at the bottom, then click “Burn into video”. You get a captioned MP4 in the original format (vertical or horizontal) that you can post as it is. It's a free alternative to the auto captions in CapCut or Canva when you don't want to install an app or create an account. Burning in re-encodes the video: expect it to take roughly the length of the video on a recent computer, longer on a phone.

SRT or VTT file for YouTube, Premiere Pro or DaVinci Resolve

If you post on YouTube, Vimeo, Facebook or LinkedIn, it's better to add subtitles as a file: viewers can turn them on or off, and search engines read their text. Download the SRT file (the most common format) or the VTT file (the web format), then upload it in YouTube Studio, under the Subtitles tab. The same SRT file imports into Premiere Pro, DaVinci Resolve, Final Cut Pro or VLC. The TXT export gives you the text alone, handy for a video description, an article or meeting notes. Before you export, read through the lines: proper names, acronyms and rare words are the most common speech recognition mistakes.

Getting the best auto subtitles

Transcription quality depends mostly on the sound: a voice close to the mic, without loud music or background noise, gives much better results. Pick the spoken language instead of automatic detection if you know it, and prefer the “Accurate” setting unless you just want a quick first try. If the subtitles show up a little too early or too late, use the global offset to move everything by a few tenths of a second. The first time you use it, the AI model is downloaded once and then kept by your browser. For videos longer than 20 minutes, a computer is best: transcribing takes time and a lot of memory, which can make a phone crash.