Pruna AI · P-Video
P-Video
The fastest way to iterate on a video idea — 5 seconds in 10
P-Video is built for speed. A 5-second 720p clip renders in about 10 seconds, and a dedicated draft tier is 4× faster still — ideal for prompt iteration, hook testing, and bulk social content. It's the only model in the Kubeez catalog with native audio-to-video lip sync: drop an .mp3 and get a talking-avatar clip whose mouth movement syncs to the waveform.
Showcase
ITERATION-FIRST VIDEO
Pruna AI's speed-optimized engine — draft mode for exploration, standard for final takes, all in one model.
FOUR TIERS, ONE TOGGLE
720p standard (7 cr/s), 720p draft (4 cr/s), 1080p standard (12 cr/s), 1080p draft (6 cr/s). Flip between them with a single click in the settings panel — or via the `quality` param on the API.
AUDIO DRIVES THE VIDEO
Upload a spoken line, a song, or a voiceover. P-Video syncs mouth movement to the waveform and uses the audio's actual duration for billing — exact to the second, not a pessimistic estimate.
DRAFT MODE — 4× FASTER
Draft is 4× faster than standard and charged at roughly half the rate. Same prompt, same input shape — the tier only changes how sharply the model renders. Explore cheap, commit once.
Automate the draft loop
Run draft by default and only escalate to standard when a quality score crosses a threshold. Both tiers share the same prompt shape, so the escalate step is a single field change in your pipeline.
Where P-Video shines
Ten hooks, one brief
Draft 720p at 4 credits per second means 10 × 5-second hooks costs you 200 credits. Test before you commit.
Audio-to-video lip sync
Drop a voice clip, get a synced clip back. No separate audio module, no manual sync pass — the video is the length of your audio.
One shot, five angles
A clean product photo becomes a rotating, panning, zooming hero clip in seconds. Standard tier for the final, draft for the iterations.
P-Video
Run P-Video in Kubeez
Open Video Studio, pick P-Video, flip Draft on, and start iterating.