Pruna AI released the p-video-2-pro video generation model

Company Pruna AI introduced the p-video-2-pro model designed for creating video content based on text descriptions or original images. The tool is already available for testing on the Replicate platform, where in the first two hours after launch, the project gained over 2200 stars.

The model supports flexible customization of generation parameters, including the choice between speed and quality modes. Users can set the video resolution in 480p or 768p formats, as well as adjust the clip duration in the range of 5 to 15 seconds. The system allows selecting various aspect ratios for adapting to different media formats.

The functionality of p-video-2-pro includes tools for managing keyframes, which allows fixing the first and last frames of the video sequence. This enables more precise control over visual composition and ensures smooth transitions when creating animations from static images.

The release expands the capabilities of generative tools on the platform, providing developers and creators with access to specialized video generation settings. Integration via API allows automating the creation of video content with specified technical parameters, simplifying the integration of the model into third-party applications and workflows.