Two original five-second AI clips, with native-language captions enabled. Caption timing follows the generated audio. Transcription and visual sampling are complete; pronunciation naturalness and precise lip sync still need native-speaker review.
Generated and checked: 12 October 2026 · local test samples
Turn on sound to listen. The CC menu also includes translated captions; changing captions does not change the audio.
お昼寝の時間です。会議はあとで。
The transcribed words match the intended line. Speech falls around 0.9–2.6 s and 3.0–4.0 s. Sampled frames show a stable face and microphone, with mouth closure at the end. Accent and individual phonemes have not been certified.
오늘은 비가 와요. 간식 챙기세요.
The transcribed words match the intended line. Speech starts around 1.2 s, with the second phrase around 3.3–4.7 s. The dog already opens its mouth and gestures during the silent lead-in, so this is not a verified precise lip-sync example. The subtitles cover the actual second phrase through its end.
Checked: complete file decoding, 1080 × 1920 video, stereo audio, no detected clipped samples, local speech transcription, and chronological frames across each clip. This does not replace listening by a native speaker.
Videos →