“That robot is going to be the best-selling product of all time. Every house gets one.”
A bulldog in a podcast booth, mid-argument on a video call. The audio, the mouth movement and the room were all produced in a single generation from a text prompt — no reference photo, and no separate lip-sync pass.
720×1280 · 8s · h264 + AAC · 24fps · 3.4 MB · opens in a new tab
+
720×1280 · 8s · h264 + AAC · 24fps · 3.3 MB · opens in a new tab
Model
google/veo-3.1-lite
@@ -98,6 +98,7 @@
Describing the breed matters more than image quality. A full-resolution character plate was reconstructed faithfully from a 150×150 thumbnail once the coat and markings were named in words.
Every line is verified, not assumed. Generated audio is transcribed back with Whisper and checked against the script before a clip is accepted.
Each shot gets a clean context. Generating several clips in one conversation caused dialogue to bleed between characters — one had another’s line appended to its own.
+
Video needs remuxing before it will stream. Model output puts the moov index at the end of the file, so a browser shows a blank player until the whole thing downloads. A lossless faststart pass moves it to the front.
Captions are burned in afterwards, not generated. Video models render text unreliably; an overlay pass gives identical typography across every clip and costs nothing to restyle.