ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds. The preview’s stated focus is more expressive character acting, camera movement, shot changes and complex visual effects; the cited launch coverage does not provide controlled...
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer

Create a landscape editorial hero image for this Studio Global article: What does ShengShu Technology’s Vidu Q4 Preview offer as the first preview of its next flagship video-generation model, as reported by IT Ho. Article summary: Vidu Q4 Preview is ShengShu Technology’s first public preview of its next flagship video-generation model, emphasizing more expressive character performances, camera storytelling and complex visual effects, with referenc. Topic tags: general, news, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers
ShengShu Technology’s Vidu Q4 Preview is the first public preview of its next flagship video-generation model. Its headline features are reference-guided creation, output options up to 4K and a launch offer starting at RMB 0.09 per second. That price is promotional and varies by configuration; the model’s advertised improvements are not, by themselves, independent proof of benchmark performance. 1
6
37
The launch coverage highlights character performance, camera language and complex visual effects. IT Home also reports improvements in dynamic camera movement, shot changes and transitions between complex shots. These describe the model’s intended strengths; the coverage cited here does not provide controlled comparisons that measure how much better it performs than other models. 6
Tencent News frames Q4 as a continuation of Vidu Q3’s production focus: Q3 emphasized keeping characters, clothing and settings consistent, while Q4’s stated priorities add greater emphasis on performance, camera expression and effects. 49
35
Vidu Q4 Preview supports two reported modes:
The model accepts up to 15 reference images for visual elements such as people, scenes and props, and up to three audio references intended to help keep character voices consistent. These inputs offer creators more ways to guide a generation, but do not guarantee that every detail will remain consistent in every result. 6
Reported output options range from 540p and 720p through 1080p, 2K and 4K. IT Home reports 10-bit color and single generations of up to 16 seconds. 6 Separate coverage also describes synchronized audio-video generation.
1
These specifications describe the available options, not a promise that every mode or price tier supports every combination. The preview currently lists image-to-video and reference-to-video modes in the launch report; check the applicable settings before planning a workflow. 6
Tencent News reports a two-month SaaS Preview promotion starting at RMB 0.09 per second. At that starting rate, a 10-second generation would cost RMB 0.90. The offer is a minimum promotional price, so it should not be assumed to apply to every resolution or configuration. 37
The promotional rate also should not be confused with API pricing: Vidu’s API documentation lists per-second credit consumption by resolution, rather than the same RMB promotional price. 2 The sources cited here do not establish the price after the promotion ends, so future costs remain uncertain.
A low starting price can make it cheaper to explore alternate prompts and shots, but the total cost of a project depends on how many generations are needed and which settings are used. The headline rate alone does not show the cost of a finished, usable sequence.
IT Home lists AI short dramas, advertising, social content and film production among the intended use cases. 6 The available launch coverage describes the model’s intended improvements, but does not provide controlled benchmark results or enough detail to verify specific scene-by-scene reviewer impressions. Treat those impressions as qualitative observations, not guaranteed behavior.
Vidu Q4 remains identified in the cited coverage as a Preview. The sources here do not confirm a date or price for a full release. 1
6
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds.
ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds. The preview’s stated focus is more expressive character acting, camera movement, shot changes and complex visual effects; the cited launch coverage does not provide controlled benchmark results.
ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds. The preview’s stated focus is more expressive character acting, camera movement, shot changes and complex visual effects; the cited launch coverage does not provide controlled...
Published byEdited with GPT-6 LunaImages generated with GPT Image 2
Research answer

Create a landscape editorial hero image for this Studio Global article: What does ShengShu Technology’s Vidu Q4 Preview offer as the first preview of its next flagship video-generation model, as reported by IT Ho. Article summary: Vidu Q4 Preview is ShengShu Technology’s first public preview of its next flagship video-generation model, emphasizing more expressive character performances, camera storytelling and complex visual effects, with referenc. Topic tags: general, news, general web. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers
ShengShu Technology’s Vidu Q4 Preview is the first public preview of its next flagship video-generation model. Its headline features are reference-guided creation, output options up to 4K and a launch offer starting at RMB 0.09 per second. That price is promotional and varies by configuration; the model’s advertised improvements are not, by themselves, independent proof of benchmark performance. 1
6
37
The launch coverage highlights character performance, camera language and complex visual effects. IT Home also reports improvements in dynamic camera movement, shot changes and transitions between complex shots. These describe the model’s intended strengths; the coverage cited here does not provide controlled comparisons that measure how much better it performs than other models. 6
Tencent News frames Q4 as a continuation of Vidu Q3’s production focus: Q3 emphasized keeping characters, clothing and settings consistent, while Q4’s stated priorities add greater emphasis on performance, camera expression and effects. 49
35
Vidu Q4 Preview supports two reported modes:
The model accepts up to 15 reference images for visual elements such as people, scenes and props, and up to three audio references intended to help keep character voices consistent. These inputs offer creators more ways to guide a generation, but do not guarantee that every detail will remain consistent in every result. 6
Reported output options range from 540p and 720p through 1080p, 2K and 4K. IT Home reports 10-bit color and single generations of up to 16 seconds. 6 Separate coverage also describes synchronized audio-video generation.
1
These specifications describe the available options, not a promise that every mode or price tier supports every combination. The preview currently lists image-to-video and reference-to-video modes in the launch report; check the applicable settings before planning a workflow. 6
Tencent News reports a two-month SaaS Preview promotion starting at RMB 0.09 per second. At that starting rate, a 10-second generation would cost RMB 0.90. The offer is a minimum promotional price, so it should not be assumed to apply to every resolution or configuration. 37
The promotional rate also should not be confused with API pricing: Vidu’s API documentation lists per-second credit consumption by resolution, rather than the same RMB promotional price. 2 The sources cited here do not establish the price after the promotion ends, so future costs remain uncertain.
A low starting price can make it cheaper to explore alternate prompts and shots, but the total cost of a project depends on how many generations are needed and which settings are used. The headline rate alone does not show the cost of a finished, usable sequence.
IT Home lists AI short dramas, advertising, social content and film production among the intended use cases. 6 The available launch coverage describes the model’s intended improvements, but does not provide controlled benchmark results or enough detail to verify specific scene-by-scene reviewer impressions. Treat those impressions as qualitative observations, not guaranteed behavior.
Vidu Q4 remains identified in the cited coverage as a Preview. The sources here do not confirm a date or price for a full release. 1
6
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds.
ShengShu’s Vidu Q4 Preview supports up to 15 reference images, three audio references and video output up to 4K, with clips up to 16 seconds. The preview’s stated focus is more expressive character acting, camera movement, shot changes and complex visual effects; the cited launch coverage does not provide controlled benchmark results.