Captio AI
🦻 Accessibility + 💻 Productivity App

Korean live captions and productivity tool for lectures

For deaf and hard of hearing people.

Download for free

Download on theApp Store
Captio AI app showing live captions

Download for free

Download on theApp Store
Use Cases

Every situation covered

Designed for the real moments that matter — not just quiet controlled environments.

🏛️Large lecture halls

Seoul National University, Yonsei, KAIST, POSTECH, and Korea University lecture halls. Korean-medium lectures at academic pace. Place the phone on the desk facing the PA system. Captio AI captions from any row.

🔬Seminars and tutorials

Korean university seminars (세미나) with students presenting and debating in fast academic Korean. Captio AI captions whoever is speaking so you follow every argument without asking for repeats.

💻Online and recorded lectures

Recorded lectures on K-MOOC, university LMS portals, or Zoom recordings. Put the recording on speaker and point Captio AI at the audio for accurate captions.

📚Study groups

Study groups in Seoul, Daejeon, or Busan — fast conversational Korean mixed with English academic terms. Captio AI follows whoever is explaining without any switching.

🌍Foreign-language and international courses

Korean students in English-medium programmes, or international students following Korean-medium lectures. Live Translator captions in real time so language is not the access barrier.

🎤Guest speakers and public talks

Guest speakers from industry and visiting academics. Captio AI handles Seoul, Gyeongsang, and Jeolla Korean — any speaker at any Korean university.

Performance

Most accurate and fastest models

Korean
95.7%

character accuracy

independent benchmark, 2025

< 200ms
you speak
captions

real-time captions

first word appears as you speak

Download for free

Download on theApp Store
Captio AI in everyday use
Why Captio AI

Built for the pace and pressure of a real lecture

Follow without splitting your attention

Lip-reading and note-taking at the same time is an impossible cognitive load. Captio AI handles the following so you can direct all your attention to understanding what is being said.

Captions while the lecturer is still speaking

Words appear on your screen as they are said — not after the sentence ends. You stay in sync with the lecture rather than always being a beat behind.

Works in any lecture setting

Lecture halls, seminar rooms, online classes, study groups. Place the phone on your desk, point it toward the audio source, and Captio AI handles the rest.

A full transcript to study from

Record the lecture and upload it to Audio Transcription after class. Get a complete written record you can search, annotate, and revise from — at your own pace.

60+ languages

For international students, students studying abroad, or courses taught in a second language. Captio AI captions or translates in real time with no manual setup.

No advance notice needed

Official university captioning requires booking in advance and is not available for every session. Captio AI works for any lecture, any time, with no prior arrangement.

Korean

Why Korean is hard to understand

3-way stop invisibility

  • Korean's three-way consonant contrast — plain, tense, aspirated — produces identical lip shapes
  • Korean distinguishes three categories for stops and affricates: plain (ㅂ, ㄷ, ㄱ, ㅈ), tense or fortis (ㅃ, ㄸ, ㄲ, ㅉ), and aspirated (ㅍ, ㅌ, ㅋ, ㅊ). All three produce the same lip closure or approach — the distinctions are entirely in subglottal pressure, laryngeal tension, and aspiration. Research on Korean lip-reading found that many consonants and vowels share identical lip movements, and that a one-to-one relation between Korean phonemes and lip shapes does not exist. In a lecture, this means every stop and affricate is visually ambiguous.

Coda invisibility

  • Korean syllable-final consonants are produced inside the mouth — the lip shape gives no information about the coda
  • Korean syllable structure allows complex codas (받침, batchim). The coda consonant is produced with the tongue, not the lips — a final /k/, /p/, or /t/ (unreleased in coda position) produces no visible movement. A deaf student lip-reading Korean hears the nucleus from the lip shape but cannot determine the coda. In a lecture, the coda carries grammatical information: 학생 (haksaeng, student) vs. 학 (hak, crane) differ only in whether a final syllable is present.

Seoul-Gyeongsang prosody gap

  • Seoul Korean and Gyeongsang Korean have different lexical pitch accents — producing different prosodic rhythms on the lips
  • Standard Seoul Korean lacks lexical pitch accent, but Gyeongsang Korean (spoken across Busan, Daegu, and much of South Korea's southeast) has a two-pattern pitch accent system. Gyeongsang-dialect Korean produces different rhythmic groupings and stress patterns from Seoul standard. Korean universities draw faculty from across the peninsula. A deaf student calibrated to Seoul Korean encounters genuinely different prosodic patterns when the lecturer grew up speaking Gyeongsang or Jeolla variety.

ㅐ/ㅔ vowel merger

  • Seoul Korean speakers have merged the ㅐ and ㅔ vowels — creating homophones that only orthography distinguishes
  • In contemporary Seoul Korean, most speakers under 60 have merged the traditional vowels ㅐ (ae, /ɛ/) and ㅔ (e, /e/) into a single vowel. Words that were previously distinct are now homophones in speech. For a deaf student, this merger compounds the difficulty: two different words share the same lip shape, and only written context distinguishes them.

Sino-Korean academic vocabulary

  • Korean university lectures use Sino-Korean (한자어) academic vocabulary with different phonological patterns from native Korean words
  • Korean academic vocabulary draws heavily on Sino-Korean (hanja-based) words: 강의 (gangeui, lecture), 연구 (yeongu, research), 철학 (cheolhak, philosophy). These Sino-Korean words have different phonological structures — consonant clusters, syllable weight — from native Korean vocabulary. A deaf student who lip-reads everyday Korean encounters different mouth shapes when the professor delivers academic content in Sino-Korean register.

Download for free

Download on theApp Store
The Challenge

Nearly 400,000 registered deaf and hard of hearing in Korea — with only 5% of special education teachers knowing sign language

The number of registered deaf and hard of hearing people in South Korea grew from 82,906 in 2000 to 384,672 in 2020, according to nationwide surveys of people with disabilities. A 2023 situational analysis of ear and hearing care in South Korea published in Frontiers in Public Health (fpubh.2023.1215556) examined the infrastructure gap, finding that despite Korea's advanced healthcare system, hearing aid adoption and intervention rates lag behind prevalence data. As of 2024, the Republic of Korea's Health Ministry reported up to 2.65 million people with some form of disability — and hearing-related disability remains among the most prevalent categories.

For deaf and hard of hearing students at Korean universities, the structural situation is constrained. A 2012 survey found that only 5% of special education teachers in Seoul knew Korean Sign Language (KSL). Korea has 2,968 registered KSL interpreters for approximately 384,672 registered deaf and hard of hearing people — a ratio of 5.7 interpreters per 100,000 population, significantly below what is required for consistent university-level provision. Oralism-focused educational approaches, where deaf students were historically taught to lip-read and speak rather than use sign language, remain influential in Korean educational policy.

The E-ASR journal study (2023) tracking deaf and hard of hearing status in Korea from 2000 to 2020 documented that while the registered population has grown, systemic support in mainstream higher education has not kept pace. Most Korean universities do not provide real-time captioning in standard lectures. Deaf and hard of hearing students must either attend Korea Polytechnic University (which has specialist DHH provision) or negotiate individual accommodation at mainstream universities — a process that varies significantly by institution and faculty.

Download for free

Download on theApp Store
Privacy

What you hear stays with you.

Gone the moment it ends

Captio AI processes your audio in real time and discards it immediately. Nothing is recorded. Nothing is kept.

Not a product

Your conversations are not something Captio AI sells, shares, or monetizes — to anyone, for any reason.

Not used to train anything

What you say in a doctor's appointment or a job interview never ends up in a training dataset. Not ours. Not anyone else's.

100% PRIVATE

Private by default

Download on theApp Store
Reviews

People love Captio AI

"SNU lectures use heavy Sino-Korean academic vocabulary. The three-way consonant system is impossible to see. Captio AI gives me full text in real time. Seminars are something I participate in now."

Jiyeon K. 🇰🇷
jiyeon.k***@gmail.com
Hard of hearing, Seoul

"My PUSAN National professor lectures in Gyeongsang-dialect Korean. My lip-reading was trained on Seoul standard. Captio AI handles both accurately. I follow the full lecture without gaps."

Hyeonsoo P. 🇰🇷
hyeonsoo.p***@gmail.com
Deaf student, Busan

"KAIST lectures mix Korean and English constantly. Captio AI switches without missing anything. I finally understand both the Korean explanation and the English term in the same sentence."

Sohee L. 🇰🇷
sohee.l***@gmail.com
Hard of hearing, Daejeon

"Jeolla-dialect Korean in my professor's delivery was different from Seoul standard. Captio AI handles both. Chonnam National University lectures are fully accessible to me."

Minjun C. 🇰🇷
minjun.c***@gmail.com
Hard of hearing, Gwangju

"SNU lectures use heavy Sino-Korean academic vocabulary. The three-way consonant system is impossible to see. Captio AI gives me full text in real time. Seminars are something I participate in now."

Jiyeon K. 🇰🇷
jiyeon.k***@gmail.com
Hard of hearing, Seoul

"My PUSAN National professor lectures in Gyeongsang-dialect Korean. My lip-reading was trained on Seoul standard. Captio AI handles both accurately. I follow the full lecture without gaps."

Hyeonsoo P. 🇰🇷
hyeonsoo.p***@gmail.com
Deaf student, Busan

"KAIST lectures mix Korean and English constantly. Captio AI switches without missing anything. I finally understand both the Korean explanation and the English term in the same sentence."

Sohee L. 🇰🇷
sohee.l***@gmail.com
Hard of hearing, Daejeon

"Jeolla-dialect Korean in my professor's delivery was different from Seoul standard. Captio AI handles both. Chonnam National University lectures are fully accessible to me."

Minjun C. 🇰🇷
minjun.c***@gmail.com
Hard of hearing, Gwangju

Download for free

Download on theApp Store
FAQ

Frequently asked questions

Follow every lecture in Korean.

Captio AI is free to start

Download for free

Download on theApp Store