Summary
ShengShu Technology has released a preview of Vidu Q4, its next flagship video model, through Vidu’s web product and API. It generates clips up to 16 seconds and offers output from 540p to 4K, with 10-bit color at the upper resolutions. Reference-to-Video accepts as many as 15 images and three short audio references, intended to preserve characters, costumes, props, locations and voices; Image-to-Video animates one starting image. The company says the model also coordinates dialogue, sound effects, camera changes and action, and improves facial performance, body motion, lighting and visual effects. Launch pricing starts at $0.014 per second, although final features, prices and availability can vary by plan and region. ShengShu says user feedback from the preview will shape the full Q4 release.
Hana’s perspective
I think the most consequential specification is not 4K but the combination of many visual references with several audio references. Production teams need repeatable characters, props, environments and voices more than they need a single spectacular shot. A 16-second ceiling still keeps Q4 Preview in shot-generation territory, so continuity across edits and reliable revision remain the harder workflow problems. The low headline price also needs caution: usable cost depends on how many attempts a team discards. ShengShu is right to expose the preview to real projects before final release; independent tests of consistency, controllability and total cost will matter more than curated examples.
Are you a writer with a great story? We'd love to hear from you: APPLY HERE