Soku AIUpload the clip
Up to 15 seconds of footage with speech in it.
All tools
Auto Subtitle Generator Captions Burned Into the Video
Upload a clip and get it back with the words on screen, highlighted one at a time as they are spoken. The captions are part of the picture, not a sidecar file — so they survive every upload, every platform and every viewer who never turns the sound on.Upload a clip and get it back with the words on screen, highlighted one at a time as they are spoken.
SubtitlesCaptions
Three deliberate steps
Say what is being spoken
The language is detected on its own; the pick decides the typeface, so Chinese, Japanese and Korean get a face that has their characters.
Get it back captioned
An MP4 with the words burned in, sized and placed for vertical placements.
Use cases
Who reaches for this, and when
Ship ads that work on mute
Most paid social is watched with the sound off. Captions are what carries the hook, the offer and the call to action to a viewer who never hears a word of it.
Caption a batch of UGC without an editor
Creator footage arrives raw. This puts the same caption treatment on every cut without opening a timeline.
Make a long take skimmable
Word-by-word highlighting gives the eye something to follow, which is most of why the format holds attention past the first second.
Questions before you run it
- Is the auto subtitle generator free?
- Yes — sign up on this page for free credits. It is the cheapest tool here to run, because nothing is generated: the clip is transcribed and the words are drawn onto the frames it already has.
- Do I get a .srt file?
- No. The captions are burned into the video, which is what ad platforms actually play. A sidecar subtitle file is dropped by most of them.
- Do I have to pick the language?
- The transcription detects it on its own, so a wrong pick will not mistranscribe anything. What the pick decides is the typeface: Latin fonts have no Chinese, Japanese or Korean characters, and a clip captioned in the wrong face comes back as a row of empty boxes.
- Does it work on Chinese, Japanese or Korean footage?
- Yes. Pick that language and the captions are set in a face that carries the characters, with a wider window — three words of English is three whole words, while three characters of Chinese is half of one.
- Where do the captions sit?
- Low in the frame but clear of the bottom strip, because Reels, Stories and TikTok paint their own caption, handle and CTA over that strip.
- What if the clip has no speech?
- There is nothing to transcribe, so nothing is drawn. Music-only footage is not what this is for.
