On September 9, 2025, Shengshu Technology made a groundbreaking announcement regarding its video large - scale model, Vidu. The company officially rolled out the Vidu Q1 reference image generation function.

This innovative function stands out as it allows for the simultaneous input of 7 reference images. In the domestic market, this breaks through the existing generation limitations. Moreover, when compared to Flux.1 Kontext, Vidu Q1 excels in crucial metrics such as consistency and fidelity, putting it on a par with and directly challenging Google's Nano Banana.

The Vidu Q1 reference image generation comes with five remarkable highlights:

• Industry - Leading Reference Quantity with Robust Controllability: It enables users to work with a large number of reference images while maintaining precise control over the generation process.
• Enhanced Subject Consistency: Whether dealing with multiple people or various scenarios, this function ensures that the subject characteristics remain stable throughout the generated content.
• High Fidelity: The generated images closely mirror the user's original input, capturing even the subtlest details.
• Extremely High Creative Freedom: Users have the flexibility to change outfits, swap backgrounds, and freely compose the images according to their creative vision.
• Support for 1080P High - Definition Resolution and Strong Chinese Semantic Understanding: It not only delivers high - quality, high - definition images but also has a deep understanding of Chinese semantics, making it more suitable for the Chinese market.