The stage is the smallest part of this announcement
Read the equipment list first, because it is ordinary. The new studio at iQIYI's Creative Center in Shanghai covers about 110 square meters. It shoots with multiple cameras, records picture and sound in sync at high resolution, keeps a digital continuity log and lets the crew review footage in real time. None of that requires AI, and none of it is new technology. A stage this size is a controlled environment for capturing performances, rather than a conventional film set built around finished scenery.
What is new is where the footage goes. It syncs automatically from the set into NadouPro, iQIYI's production platform, where the release says filmmakers get access to nearly 30 AI models, naming Seedance 2.5 and Kling 4.0, alongside hundreds of production-focused AI agents and skills. The studio hardware is the input device. The product iQIYI is actually announcing is the pipe between a performance and a shelf of video models, with the transfer step removed.
Performance first, the set afterwards
That pipeline puts the emphasis somewhere different from prompt-to-video generation. Instead of asking a model to invent the actor and environment together, iQIYI starts with a recorded human performance and uses AI to build or adjust the surrounding scene. Actors perform the scene, including action sequences staged against a plain backdrop, and those recordings become the reference for generation: appearance, performance, composition, lighting and spatial relationships are captured, and the environment is built around them afterwards.
The release's before-and-after still illustrates the pitch more plainly than its prose. The upper image shows an actor in modern dress performing in front of a white curtain under studio lights. The lower image presents what iQIYI describes as the corresponding performance placed inside a forest environment. It is a company-supplied demonstration of the intended workflow, not independent evidence of how consistently the process works across a finished production. Liang Zhu, the company's head of intelligent production, describes the objective as preserving the performance while adjusting sets, lighting, costumes, makeup and props afterwards.
The numbers that are not in the release
Three cautions before this gets quoted back as a turning point. First, "China's first AI film studio" is iQIYI's description of its own facility. The release names no comparison set and no one else certifies the ranking; treat it as branding until a competitor or a trade body contests or confirms it. Second, the one efficiency number in the release, gains of over 30%, belongs to the company's earlier Hengdian virtual studio. It is evidence that iQIYI has run virtual production before. It says nothing measured about this room, and the release provides no baseline or method for it in any case.
Third, the release does not provide the figures needed to judge the production impact. It gives no share of AI-generated footage in "Cop in Trouble," no cost per finished minute, no comparison of production timelines against a conventional shoot, and no release date for the film. These are not minor omissions. Without them, readers cannot determine whether the studio materially changes production economics or mainly offers a more integrated post-production workflow. The November 2 announcement establishes a new pipeline and a first production, not a proven improvement in filmmaking efficiency.
For filmmakers outside China, the practical significance is integration rather than exclusive model access. iQIYI says NadouPro combines nearly 30 AI models with hundreds of production-focused agents and skills, while the studio supplies standardized performance footage directly to that platform. Whether this combination offers a meaningful advantage over existing production workflows remains unproven. The company has not published throughput, cost or production-time comparisons for the new facility.
Sources
- [1] iQIYI Inc. press release — "iQIYI Launches China's First AI Film Studio in Shanghai, with Inaugural Title 'Cop in Trouble' in Production," November 2, 2026, via PR Newswire (announcement date and October 28 filming start; Shanghai Creative Center location; approximately 110 square meters; multi-camera shooting, synchronized high-resolution video and audio, digital continuity logging, real-time review; automatic sync to NadouPro; nearly 30 AI models including Seedance 2.5 and Kling 4.0; hundreds of production-focused AI agents and skills; Liang Zhu quote; "Cop in Trouble" plot; Hengdian virtual studio efficiency claim of over 30%, which refers to the earlier facility and carries no baseline or method; read first-hand)Read source
- [2] AI Weekly — "iQIYI Opens China's First AI Film Studio in Shanghai, Starts Shooting Internet Feature 'Cop in Trouble'," November 2, 2026 (independent alert; used only for timing corroboration, no facts taken from it)Read source