Stop pausing to look up words. Hover any word for instant definitions and pronunciation — on YouTube, Netflix, Twitch, or any website. No subtitles required.
Start with 30 free minutes. No signup required.
Works on every website
Every language works with real-time audio transcription — even when there are no subtitles.
Install the extension, play a video, and start learning. That's it.
YouTube, Netflix, Twitch, Bilibili, TikTok — any website with video or audio content.
Real-time transcription starts instantly. If subtitles exist, we use them. If not, we transcribe the audio live.
Hover any word for pronunciation and definition. Save words to build your vocabulary. Learn naturally through immersion.
Built for language learners who want to understand native-level videos, not textbooks.
AI-powered speech-to-text that works on any video, even without subtitles. Hear something? We'll transcribe it.
Automatically detects existing subtitle tracks and uses them when available — free, instant, and more accurate.
Hover any word to see pronunciation, pinyin, romaji, and instant English definitions. No copy-pasting into Google Translate.
Save words with one click while watching. Review your saved vocabulary anytime. Export to Anki or other SRS tools.
Hear native speakers pronounce any word instantly. Tap a word, hear it spoken — across all 15 supported languages.
Beautiful stacked subtitles with adjustable size and opacity. Feels like watching a movie, not using a tool.
Every competitor depends on subtitle tracks. We transcribe the audio.
|
golingu
|
Language Reactor | Trancy | Migaku | |
|---|---|---|---|---|
| Works without subtitles | ✓ Audio transcription | ✗ | ✗ | ✗ |
| Works on any website | ✓ Every site | YouTube + Netflix | Limited sites | Limited sites |
| Word-level lookup | ✓ | ✓ | ✓ | ✓ |
| Vocabulary saving | ✓ | Pro only | ✓ | ✓ |
| Live streams | ✓ | ✗ | ✗ | ✗ |
| 15 languages | ✓ | Many | Some | 3 |
| Price | $7/mo (audio + subs) | $5/mo (subs only) | $8/mo (subs only) | $9-14/mo (subs only) |
Start free. Upgrade when you need more minutes.
Cancel anytime. 14-day money-back guarantee.
Join thousands of language learners using golingu every day.
"I've been studying Mandarin for 2 years and this is the best tool I've found. Being able to hover over words in C-drama dialogues and instantly see pinyin + definitions is incredible. Way better than pausing to look things up."
"The fact that it works on Bilibili without needing subtitle files blew my mind. I watch a lot of Chinese content that has no English subs, and this just transcribes the audio live. Game changer for immersion learning."
"I switched from Language Reactor because half the videos I watch don't have subtitle tracks. golingu works everywhere — even on random livestreams. The vocab saving feature helps me remember what I learn too."
Everything you need to know about golingu.
golingu captures the audio from your browser tab and sends it to Deepgram's nova-2 AI for real-time speech-to-text transcription. This means it works on any video, even livestreams or sites with no subtitle tracks at all. If subtitles are available, we detect and use those automatically (free, instant, more accurate).
Every website. YouTube, Netflix, Bilibili, TikTok, Twitch, Disney+, Crunchyroll, news sites, course platforms — any page that plays video or audio in your browser tab. If you can hear it, golingu can transcribe it.
We support 15 languages: Chinese (Mandarin), Japanese, Korean, Spanish, French, German, Portuguese, Italian, Russian, Arabic, Hindi, Vietnamese, Turkish, Indonesian, and English. More languages are being added regularly.
Free gives you 30 minutes/month of audio transcription and unlimited use of subtitle-based mode. Pro gives you 3,000 minutes/month (~50 hours) of audio transcription, plus vocabulary export and priority support. Both tiers include word-level lookup, vocab saving, and all 15 languages.
The browser extension works on desktop Chrome and Microsoft Edge. A companion mobile app for vocabulary review and flashcards is planned for a future release.
We use Deepgram nova-2, one of the most accurate real-time speech-to-text models available. Accuracy is typically 90%+ for clear audio. Results can vary with background noise, accents, and audio quality, but it's more than good enough for learning.
Install in 10 seconds. No signup required.
Get Free Chrome ExtensionAlso available for Microsoft Edge