AI Video Captions
Add trending animated captions to any video. 6 styles, word-level animation, multi-language support. Self-hosted and free.
Drop your video here or click to browse
MP4, MOV, or WebM up to 500MB
Leave on auto-detect unless the transcript comes out in the wrong language.
Everything You Need
A complete toolkit for adding stunning animated captions to any video.
AI Transcription
Powered by faster-whisper. Automatic speech-to-text with word-level timestamps for precise caption timing.
6 Caption Styles
Hormozi, MrBeast, Karaoke, Minimal, Bounce, Classic. Trending styles from top creators.
Word-Level Animation
Each word animates individually — highlights, karaoke wipes, bounces, and scale effects.
Multi-Language
Automatically detects 100+ languages. Script-aware font selection for Latin, CJK, Cyrillic, Arabic, and more.
Self-Hosted & Private
Run on your own infrastructure with Docker. No accounts, no tracking, no data collection.
HD Export
Download your captioned video in full quality. CRF 18 encoding, audio preserved.
How It Works
Three simple steps from raw video to captioned masterpiece.
Upload Video
Drop any MP4, MOV, or WebM video. Up to 500MB and 30 minutes.
Pick a Style
Choose from 6 trending caption styles. Preview the look in real-time with our interactive phone mockup.
Download
Your captioned video is ready. Download in full HD quality.
Frequently Asked Questions
Everything you need to know about AI Video Captions.
We support MP4, MOV, and WebM video files. Most videos from phones, cameras, and screen recorders work out of the box.
It depends on video length and your hardware. Roughly 1-3 minutes per minute of video on a modern CPU. Shorter videos process in seconds.
Over 100 languages are supported via automatic detection. The AI transcription identifies the spoken language and selects appropriate fonts for each script.
Yes! Use our interactive phone mockup to drag captions anywhere from 5% to 50% from the bottom. Preset buttons for Top, Middle, and Bottom positions.
Default limits are 500MB and 30 minutes, but these are configurable via environment variables when self-hosting. Adjust them to match your server's capabilities.
It's fully self-hosted and open-source. Your videos stay on your infrastructure. No accounts, no uploads to third parties, no usage limits, and no vendor lock-in.
100% Open Source
No accounts. No data collection. No limits. Self-host with Docker Compose in one command. MIT licensed.
$ git clone https://github.com/nicolaigaina/
ai-video-captions.git
$ cd ai-video-captions
$ docker compose up
# Open http://localhost:3000