Automatic captions
Automatic video caption generator
Drop in a short video and get captions timed to every word, in minutes, free and in your browser. Pick one of 150 animated looks, fix any word it heard wrong, and export an MP4 ready to post or a subtitle file for YouTube. Nothing to install, no watermark, and for transcription and the normal MP4 export your video file is never uploaded anywhere.
How it works
1. Upload your video
Open Captionit and drop in a clip. MP4, MOV, WebM and MKV work. The file stays on your device.
2. Transcribe in your browser
Whisper runs inside the tab and writes out what was said, timed to every word, in one of 14 languages including Hindi and Hinglish. If you only need the words, the standalone Transcribe tool does the same job and hands you a timed transcript.
3. Pick a look
Start from one of 150 animated caption looks, then change the font, colours, position and motion to match your video. Add a subscribe pill, a counter or a title from Motion Studio when the video needs one.
4. Fix any word
Click a word and type the correction. The timing stays. If the whole track is early or late, shift all captions at once instead of dragging them one by one.
5. Export and post
Export an MP4 with the captions burned in, ready for Reels, Shorts and TikTok, or a subtitle file (SRT, VTT, ASS, plain text or JSON) for YouTube and other editors.
What you get
- Word-timed captions. Every word lands on the frame where it is spoken, and the key word on each line gets its own colour, box or motion.
- 150 animated looks, the largest style count we verified among rival tools in this research, and every look is in the free plan.
- 14 languages, including Hindi and Hinglish. Hindi and Hinglish come out in Roman script, the way most creators type; Devanagari Hindi is a separate option.
- MP4 and subtitle export from one workflow. Most tools do one or the other; here you get both.
- No watermark, nothing to install. Transcription, editing and MP4 export all run in your browser, on desktop or phone.
- Built for vertical video. 9:16 Reels, Shorts and TikTok first, with 16:9 widescreen working the same way.
Captions in your language
English, Hindi, Hinglish, Tamil, Telugu, Bengali, Marathi, Punjabi, Urdu, Spanish, French, German, Arabic and Portuguese. Hinglish captions use a Whisper model fine-tuned for Hindi-English speech, so mixed Hindi-English talking comes out as the Roman-script Hinglish creators actually post, not as a translation.
Export formats
MP4 with captions burned in; transparent overlays as MOV (ProRes) or WebM, rendered by the free desktop Companion app; and caption files as SRT, VTT, ASS, plain text or JSON. Translated tracks add translated SRT and VTT rows. MP4 export needs Chrome, Edge or Safari.
What it cannot do
It transcribes what was actually said, so it cannot write captions for silent video, and it will make mistakes on heavy background noise, overlapping speakers or very quiet speech. Those are fixed by editing the words, not by re-running. It is built for short video: free videos can be up to 5 minutes long, and paid plans allow longer videos up to 3 hours. Transparent overlay exports need the Companion app; the normal MP4 export does not.
Free, with paid passes when you outgrow it
The free plan includes the full editor, all 150 looks and every AI tool. Paid plans are 30-day passes that raise the limits: more projects, longer videos, bigger uploads, longer export storage. You pay once; nothing renews or charges you again automatically. See Pricing for the plans in your currency.
Frequently asked questions
Is the automatic caption generator free?
Yes. Captions, every look, the full editor and MP4 and SRT exports are free, with no watermark. Free videos can be up to 5 minutes long. Paid 30-day passes raise the limits — see Pricing.
Do I have to upload my video anywhere?
For transcription and the normal MP4 export, no. Speech recognition and rendering run inside your browser, so the video file is not sent anywhere for those. The file name, size, length and format are stored with your project, and your captions and transcripts are saved so you can reopen the project later. One exception: transparent overlay exports are rendered by the optional desktop Companion app, and that render does use the video from your project.
Which languages can it caption?
14 languages: English, Hindi, Hinglish, Tamil, Telugu, Bengali, Marathi, Punjabi, Urdu, Spanish, French, German, Arabic and Portuguese. Hindi and Hinglish captions come out in Roman script, the way most creators type; Hindi in Devanagari script is a separate option.
Can I fix a word it heard wrong?
Yes. Every word is timed to the speech and every word is editable. Click a word, type the correction, and the timing stays. You can also shift all captions at once if the whole track is early or late.
What can I export?
An MP4 with the captions burned in, ready to post on Reels, Shorts and TikTok; transparent overlay videos (MOV and WebM) rendered by the free desktop Companion app; and caption files (SRT, VTT, ASS, plain text and JSON) for YouTube, LinkedIn, Premiere, CapCut and other editors. MP4 export needs Chrome, Edge or Safari.
What can it not do?
It transcribes what was actually said, so it cannot invent captions for silent video, and it will make mistakes on heavy background noise, overlapping speakers or very quiet speech. Fix those by editing the words. It is built for short video: free videos can be up to 5 minutes long, and the longest paid plan allows videos up to 3 hours.
Caption your next video free
Captionit is made by one person, for creators who post every day. Read more on the About page, or go straight to the homepage to see the full studio.
Caption a video free