Дубляж пока не включён. Уже сейчас можно осмотреться и создать аккаунт.

Результаты измерений

We have no customers yet, so instead of logos: what we measured, on which clips, and when. Including what does not work well. Test setup: one laptop GPU (RTX 4060, 8 GB). Full log in the repository.

Test data

FLEURS, read sentences in Korean and 20 more languagespublic dataset (CC BY 4.0), with reference transcripts and translations
KsponSpeech, 150 clips of conversational Korean (14 min)research dataset with transcripts — local test only
529 Korean–English sentence pairs: TED talks, everyday expressions, FLEURShuman translations (OPUS TED2020, AI Hub, FLEURS) — local test only
CREMA-D, 160 acted English sentencesemotion labels, for the emotion model
A two-person dialogue, 36 smade with two synthetic voices — no real people
TV clip, Korean, 1:30 · web drama, Korean, 25 min · animated lesson, English, 5:26copyrighted — local tests only, never published

What works

ПроверкаResultDate
Speech recognition, Korean9.1 % character error rate on 22 minutes of read speech (one pass, like a real video); 11.5 % on conversation2026-10-16
Translation now: MADLAD-400 3B (allows commercial use), 100 FLEURS sentences per directionchrF++ 56.2 Korean → English, 57.2 Uzbek → English, 42.0 English → Uzbek; on a news clip 3 из 5 meaning traps right (NLLB 0 of 5; Qwen3, the Korean ↔ English second opinion, 5 of 5)2026-11-11
Translation before (NLLB-200 3.3B, non-commercial), 529 sentence pairschrF++ 52.3 Korean → English and 32.8 English → Korean (49.3 and 29.4 with the smaller model it replaced); fixed "Newt Churchill" → Gingrich, "below sea level" → below freezing2026-10-16
Говорящие различаютсяtwo synthetic voices, a man and a woman: 13 из 13 sentences right; FLEURS readers: the two groups matched the dataset's male / female labels in 15 of 15 sentences2026-10-16
A speaker's own voice, cloned from the video with consentspeaker similarity to the person 0.85 (built-in voice: 0.66)2026-10-15
Korean voices26: the built-in voice and 25 English presets speaking Korean — Whisper understood 20 of them as well as the built-in voice; a two-person Korean dub, one voice per speaker: 0.6 % character errors2026-10-16
Voice running on after the sentenceFrench dub, two sentences ended in made-up words: 11.0 % → 2.1 % character errors after DUBBE cuts at the sentence's last word; the af_alloy voice in Korean 29.3 % → 8.1 %, so it is back in every language2026-10-19
Timing drift, three clipslargest delay 0.00 с (the lesson needed the picture slow-down option; 0.37 s without)2026-10-06
Long translations (lesson)strongest voice speed-up 1.30× → 1.18×; sentences that did not fit 16 → 02026-10-06
Background sounda phone ringing kept its level (−0.0 dB); 0 words of the original voice left over2026-10-06
Voice choice by pitchmatched the speaker's gender for 95 % of Korean and 95 % of English test speakers2026-10-06
Speed29 с of processing per video minute into English (2:43 clip); 77 с into Korean (5:26 clip)2026-10-16

The other 20 languages

FLEURS read speech only — no conversation, no native listener yet. Recognition: 20 sentences in one pass, with the Whisper model that made fewer errors in that language. Translation: 100 sentences, chrF++ (higher is better). Voice: an English preset speaking 6 sentences, Whisper listening back.

ЯзыкRecognition errorsTranslation from Englishinto EnglishVoice errors
Арабский4.4 %53652.2 %
Китайский8.8 %23*556.9 %
Датский3.2 %68712.6 %
Нидерландский3.2 %54580.2 %
Финский4.5 %52601.4 %
Французский4.1 %70680.7 %
Немецкий1.8 %63702.5 %
Греческий3.2 %52626.0 %
Хинди13.0 %58676.0 %
Итальянский1.4 %59640.0 %
Японский6.8 %24*569.6 %
Малайский10.1 %66682.9 %
Норвежский3.6 %57653.1 %
Польский2.4 %47575.4 %
Португальский2.2 %69724.4 %
Русский3.8 %56641.9 %
Испанский2.4 %55603.7 %
Суахили26.3 %60676.6 %
Шведский2.9 %68720.6 %
Турецкий17.2 %58654.8 %

* chrF++ is not comparable for Japanese and Chinese (no spaces between words). Hebrew was left out: recognition 29 % and voices 38–47 % character errors. French, Hindi and Swahili use only the English presets — their built-in voice failed.

Where errors come from (FLEURS, Korean → English, with the earlier translation model)

We replaced one step at a time with a correct human reference:

What does not work well yet