Auto Captions for Video & Audio

Add automatic subtitles to web video and audio. Pick a language and get speech-to-text captions generated on your device.

Add automatic subtitles to video and audio on the web. This extension puts a small CC button on HTML5 video and audio players. Choose a language from the menu and the extension listens to the audio and shows the words as captions. It works on videos and audio that have no subtitles or transcripts of their own. How to use 1. Open a page with a video or audio player and hover over it (audio players show the button right away). 2. Click the CC button next to it. 3. Choose a language. The first time you use a language, a speech model is downloaded (this can take a moment and can be tens of megabytes). After that it is stored in your browser and reused. 4. Press play. Captions follow playback, including when you pause or jump to another point. Your audio stays on your device Speech recognition runs locally in your browser using the open-source Vosk engine (through the voice2text library). Video audio is not uploaded to any server. The only network request the extension makes is downloading the language model the first time you pick a language. Supported languages English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Ukrainian, Polish, Czech, Turkish, Arabic, Persian, Hindi, Chinese, Japanese, Korean, Vietnamese, Catalan, Kazakh, Uzbek, Breton, Esperanto, Greek, Swedish, Gujarati, Tajik, Telugu, Kyrgyz, Georgian, Tagalog. Good to know (limits) - Accuracy depends on the language model and the audio. The models are small, so expect mistakes, especially with music, background noise or several people talking. - Some sites cannot be captioned. If a site serves its video or audio from another domain without permission, or uses protected/streaming media, the browser blocks access to the audio and the extension will tell you so instead of failing silently. - Captions are generated as the media plays, so they appear slightly after the words are spoken. - Audio players: a visible player gets its own captions and button attached just below it. A hidden audio element (common on podcast/radio sites with a custom player) gets a small floating control once it has been playing for a few seconds, so short sound effects never trigger it. Open source components: Vosk (Apache 2.0) and voice2text (MIT).