Vidu Q4 Preview unveiled: Shengshu announces 4K video and reference audio

Shengshu Technology unveiled Vidu Q4 Preview, the latest version in its Vidu AI video generation family, in an announcement from Beijing on 7 October 2026. The company said the model was now available and was aimed at short-form dramas, advertising, social content and film teams. The announcement also cited goals of faster generation and lower costs; these performance and cost claims are the manufacturer’s own statements.

What features did the company announce?

Shengshu’s announcement highlights reference audio support, which is intended to help keep character voices consistent. The company says the feature supports natural emotional delivery and character performances. On the visual side, it lists 4K resolution and 10-bit color as features geared toward professional production workflows. These capabilities are described in the company’s product announcement; it does not include independent tests or comparative measurements.

According to the announcement, the model aims to deliver more natural texture and depth, improved lighting, and camera movements and cuts that respond to direction. Shengshu also says up to 15 reference images can be used together to build complex scenes and ideas. This may offer another control option for teams looking to generate video from visual references.

Pricing and promotion

The company says video generation starts at $0.045 per second at 540P resolution. Pricing increases with resolution, reaching up to $0.39 per second for 4K generation. When estimating production costs, both clip duration and the selected resolution need to be taken into account.

Shengshu announced a 30% discount on image-to-video and reference-to-video generation with Vidu Q4 Preview through 30 November. The announcement does not specify the year the promotion ends. It also does not say whether other types of generation are included, so users should check the current terms on the platform.

Implications for architecture and visualization workflows

Vidu Q4 Preview is not presented as architectural modeling or architectural visualization software. Even so, it could be worth testing as a video generation tool for architecture and interior design firms and archviz teams producing project showcases, social media content or presentation videos. Reference images may help guide the visual direction of a scene; however, the source provides no information about consistency in architectural content or integration with specific 3D software.

The announcement also does not detail how 4K and 10-bit output can be incorporated into existing workflows. Information about file formats, export options, hardware requirements and licensing terms is limited. Teams should therefore run short tests before incorporating the tool into production, and separately check whether the outputs meet delivery requirements and review the terms of use.

How should production claims be assessed?

Shengshu highlights promises of faster generation and lower costs, but the announcement includes no numerical comparison measuring speed gains or cost differences. These claims should therefore be read as the company’s product positioning, rather than as a verified performance comparison. Although pricing is given by resolution, production duration and promotional terms must also be factored in to estimate the total project cost.

Vidu Q4 Preview is said to be available. However, the announcement does not explain the scope of the usage license, terms for commercial projects or compatibility with different software. It is important to clarify these details, especially for outputs intended for client work.