Breaking

Pruna AI launches P-Video-Edit for prompt-based video-to-video changes

September 3, 2026 at 14:03 EDT

Media Generation
Foundation Models

Pruna AI has launched P-Video-Edit, an instruction-based video-to-video model that rewrites existing clips from a text prompt, with optional guidance from up to four reference images, and listed it on Design Arena as well as partner inference APIs. The company positions the tool as a way to change colorways, products, environments, objects, or on-screen text without reshooting, on source clips of up to 15 seconds.

Sep 3, 2026 · Pruna AI

Don’t reshoot. Edit the take.

P-Video-Edit applies a text instruction—and up to four reference images—to existing clips of up to 15 seconds, with a cheaper Draft preview before a Standard render.

Draft vs Standard price per second

$0.025

Draft / s

$0.045

Standard / s

Draft is billed lower for faster previews; Standard is the production-quality path. A 15s Full clip is listed at $0.675.

15s

Maximum source clip

Optional reference images

23.18s

Listed 720p inference

Edit loop

1. Source
MP4, up to 15s

→

2. Instruct
Prompt + up to 4 stills

→

3. Draft
$0.025/s preview

→

4. Standard
$0.045/s final

WHERE IT HELPS

One approved take can become color, product, style, or environment variants for ads and social clips without a full reshoot.

LIMITS

Hard 15-second cap; Draft is lower quality; some partner endpoints do not expose resolution, FPS, or format controls. Arena rankings are still pending.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.