Descriptions:
Youri van Hofwegen presents a workflow for generating AI videos up to 10 minutes long in a single generation pass, using InVideo as the production platform backed by Google’s VEO 3.1 and Sora 2 as the underlying video models. The video directly addresses what he calls the core bottleneck in AI video production: the 8-to-12-second clip limit that forces creators to manually stitch together dozens of short segments and fight consistency issues between them.
InVideo’s pre-built workflow templates — covering categories like faceless explainer videos, educational content, listicle videos, and script-to-video — handle prompt engineering automatically. The demo builds a 3-minute faceless explainer about daily productivity habits, selecting a clear American female voice, bold pop-effect subtitles for retention, and calm instrumental music. The ‘only generated clips’ setting ensures all footage is AI-created rather than drawn from InVideo’s stock library. After initial setup, the media generation step unlocks model selection via the Pro tier, where VEO 3.1 and Sora 2 drive the final visual quality.
The resulting video uses fully AI-generated footage throughout, with the creator noting the output quality is noticeably stronger than typical AI video tools at this length. The workflow is positioned as a significant time-saver for YouTube creators, educators, and marketers producing long-form content without access to traditional production resources.
📺 Source: Youri van Hofwegen · Published December 13, 2025
🏷️ Format: Tutorial Demo
