Breaking

Pruna Launches p-video-2-pro-reference With Full Multimodal Control on Design Arena

October 15, 2026 at 15:08 EDT

Foundation Models
Media Generation

Pruna AI has released p-video-2-pro-reference on Design Arena, expanding the capabilities of its P-Video-2-Pro model with comprehensive image, video, and audio reference conditioning. Built on MiniMax H3, the new reference variant enables creators to exercise granular control over video generation by providing multiple reference materials that guide character consistency, motion behavior, and audio design in a single workflow.

October 15, 2026 • Pruna AI

Pruna Adds Multimodal Reference Control to P-Video-2-Pro

p-video-2-pro-reference now accepts images, videos, and audio to guide character consistency, motion transfer, and audio design in a single generation.

Reference Capacity

9 Images

up to 20 MiB combined

Video References

3 Videos

up to 100 MiB total

Audio References

3 Files

up to 15s combined

Generation Speed

4.3s

for 5s at 768p

Key Capabilities

Character consistency across multiple shots via reference images
Motion transfer from reference videos to target characters
Native stereo audio generation synchronized with video
Soft first-frame and last-frame keyframe control
Aspect ratio adaptation to match reference proportions

Use Cases

Advertising and branded content where character and voice consistency are critical; animation workflows requiring motion-transfer control; e-commerce product videos with character talent; UI/UX animation with consistent visual language across scenes.

Continue reading

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