Local Live Captions Autor: Eric
Live English captions for any audio or video in a tab — video, podcasts, meetings, streams. Speech recognition runs locally in your browser with Whisper; your audio never leaves your computer.
Metadane rozszerzenia
O tym rozszerzeniu
Local Live Captions puts a caption panel over whatever is playing in a tab, the way Chrome's Live
Caption does — except the speech recognition runs on your own machine, inside Firefox,
using a Whisper model compiled to WebAssembly.
• Captions appear on their own as soon as a tab starts playing audio
• Everything is local: after the one-time model download the add-on works offline,
and your audio is never uploaded
• Works with ordinary <video>/<audio> players, with sites that play through the Web
Audio API, and with a microphone source for calls or anything Firefox cannot tap
• Drag the panel anywhere, resize it, change the text size, pick dark or light
• Choose your accuracy/speed trade-off: whisper-tiny.en for the fastest captions,
base or small when you want more accuracy; WebGPU if your build supports it
• Can translate other languages into English with a multilingual model
• Optionally point it at your own transcription server instead (OpenAI-compatible)
Keyboard shortcut: Ctrl+Shift+L (Cmd+Shift+L on macOS) toggles captions for the tab.
First run downloads the speech model (about 40 MB for the default). Firefox asks you
to grant site access before the add-on can read a page's audio.
Known limits: media served cross-origin without CORS headers cannot be tapped (use
microphone mode there), DRM video cannot be captured at all, and captions lag by about
the length of the phrase being spoken.
Caption does — except the speech recognition runs on your own machine, inside Firefox,
using a Whisper model compiled to WebAssembly.
• Captions appear on their own as soon as a tab starts playing audio
• Everything is local: after the one-time model download the add-on works offline,
and your audio is never uploaded
• Works with ordinary <video>/<audio> players, with sites that play through the Web
Audio API, and with a microphone source for calls or anything Firefox cannot tap
• Drag the panel anywhere, resize it, change the text size, pick dark or light
• Choose your accuracy/speed trade-off: whisper-tiny.en for the fastest captions,
base or small when you want more accuracy; WebGPU if your build supports it
• Can translate other languages into English with a multilingual model
• Optionally point it at your own transcription server instead (OpenAI-compatible)
Keyboard shortcut: Ctrl+Shift+L (Cmd+Shift+L on macOS) toggles captions for the tab.
First run downloads the speech model (about 40 MB for the default). Firefox asks you
to grant site access before the add-on can read a page's audio.
Known limits: media served cross-origin without CORS headers cannot be tapped (use
microphone mode there), DRM video cannot be captured at all, and captions lag by about
the length of the phrase being spoken.
Ocenione na 0 przez 0 recenzentów
Uprawnienia i dane
Wymagane uprawnienia:
- Mieć dostęp do kart przeglądarki
- Mieć dostęp do danych użytkownika na wszystkich stronach
Opcjonalne uprawnienia:
- Mieć dostęp do danych użytkownika na stronie „huggingface.co”
- Mieć dostęp do danych użytkownika na stronie „cdn-lfs.huggingface.co”
- Mieć dostęp do danych użytkownika na stronie „cdn-lfs-us-1.huggingface.co”
Zbieranie danych:
- Autorzy tego rozszerzenia twierdzą, że nie wymaga ono zbierania danych.
Więcej informacji
- Odnośniki dodatku
- Wersja
- 1.3.0
- Rozmiar
- 5,18 MB
- Ostatnia aktualizacja
- 20 godzin temu (16 wrz 2026)
- Powiązane kategorie
- Licencja
- Licencja MIT
- Prywatność
- Zasady ochrony prywatności tego dodatku
- Historia wersji
- Dodaj do kolekcji