Watch anything.
Understand everything.

Real-time transcription & translation captions
for any streaming media — right in your browser,
That works.

Any website Tone-aware Near-instant latency 100+ languages
youtube.com/watch?v=…

Intelligence

Translates tone,
not just words.

Typingo detects the register of what's being said — casual, polite, formal, or narrative — and adapts the translation to match. A vlog translates differently from a lecture.

Each sentence also carries the context of the previous one, so terminology, pronouns, and references stay consistent across the whole session.

Casual 해체 · 반말
Polite 해요체
Formal 합니다체
Narrative 하다체

Quality mode · English → Korean

Can we measure it with an awesomometer?
Quality '멋짐 측정기' 같은 걸로 그 멋짐을 재볼 수 있을까요?
Made-up word "awesomometer"coined Korean equivalent · not a transliteration
That is artificial intelligence, which is seemingly everywhere over the past few years.
Quality 여기서 '코끼리'는 바로 인공지능(AI)을 말해요. 최근 몇 년 동안 AI는 정말 어디에나 있는 것 같아요.
References previous sentence's "elephant" metaphor — cross-sentence context
Harvard University's introduction to the intellectual enterprises of computer science.
Quality 하버드대학교에서 제공하는 컴퓨터 과학 입문 과정 CS50이에요.
Natural Korean phrasing · concise · informal register auto-detected

Speed

Near-instant translation.
In half a second.

The first translated word appears in as little as ~216 ms — often before the sentence has even finished. By the time the speaker takes their next breath, the full translation is already on screen.

Audio-video sync delay lets the translation pipeline run ahead of playback, so captions are never playing catch-up.

Translation latency · English → Korean

First token
~50 ms
Avg TTFT
~216 ms
Avg · Speed
~370 ms
Avg · Quality
~585 ms
P90 · Quality
~720 ms
Speed mode Quality mode

Compatibility

Works on any website.

Typingo captures audio at the browser-tab level — not from specific video players or platform APIs. If it plays in Chrome, Typingo can translate it.

YouTube · Netflix · Zoom · Twitch · Coursera · Google Meet · Vimeo · X / Twitter · TED · Any website
No platform integrations required. No DOM access needed. Works via Chrome's tab audio API.

Diarization

Beta

Knows who's
speaking.

In interviews, panels, and group calls, speakers get their own caption track. Each voice is separated, translated independently, and displayed in its own color so you always know who said what.

Supports up to 10 simultaneous speakers. Currently in beta.

Live caption overlay ● Live
Speaker 1
이 기술은 앞으로 어떻게 발전할 것 같으신가요?
How do you think this technology will develop going forward?
Speaker 2
글쎄요, 저는 특히 실시간 번역 분야에서...
Well, I think especially in the real-time translation space...