Alibaba-backed PixVerse launches real-time AI video tool

Beijing — Anne alibabaA startup backed by is pushing artificial intelligence towards real-time interactive video creation.

PixVerse on Tuesday released an AI tool that allows users to control how videos unfold during generation. Also, just like a movie director, users can instruct their characters to cry, dance, pose, etc., and the action will happen instantly while the video plays.

Real-time AI video generation can create “new business models,” co-founder Jaden Xie told CNBC in an interview translated from Chinese. Possibilities include a world where users can influence the development of microdramas or play “open-ended” video games that aren’t constrained by pre-designed storylines, he said.

Founded in 2023, PixVerse raised more than $60 million in the fall, with Alibaba leading the round and Antler also participating.

Mr. Xie did not disclose the amount, but said the company was nearing completion of its next funding round. According to him, more than half of the participating investors are from overseas.

PixVerse’s latest AI tools highlight how a China-based team is giving competitors an advantage with AI-generated video tools.

With the exception of Israeli startup Lightricks, the top eight AI video generation models tracked by AI benchmarking firm Artificial Analysis are Chinese companies. Many offer faster generation speeds and much lower usage fees than OpenAI’s advanced premium AI video model, Sora 2 Pro.

OpenAI’s Sora first gained global attention nearly two years ago when it unveiled a new text-to-video generative model, but it didn’t become publicly available until December 2024. By that time, several Chinese teams had already released competing tools to users around the world.

“Sora still defines the upper limit of quality in video generation, but is limited by generation time and API costs,” said Wei Sun, principal analyst at market research firm Counterpoint, referring to how users are charged access fees for using AI models.

“Chinese companies are taking a different path. They are turning video generation into a scalable, low-cost, high-throughput production tool.”

Just last month, Beijing-based startup Shengshu announced that its TurboDiffusion video framework, developed with researchers at Tsinghua University, can create videos 100 to 200 times faster with minimal quality loss.

PixVerse’s latest AI tools eliminate most of the latency and aim to be more than just special effects. The tool is embedded in the company’s social media-style sharing platform, which surpassed 16 million monthly active users in October.

Real-time video generation bridges the gap between content creation and delivery, reshaping the way users interact with AI-generated content, Xie said.

Xie aims to reach 200 million registered users in the first half of this year, up from 100 million in August, and roughly double his team to nearly 200 people by the end of the year.

PixVerse primarily serves users outside of China through a web browser interface and smartphone app.

Alyssa Lee, DataHub’s chief of staff and former vice president at Bessemer Venture Partners, said that compared to Chinese-made AI video tools, “most of the U.S. products are relatively simple and minimal” in user interface and experience.

Lee said scenario-specific AI video tools offer a clearer path to monetization, noting: adobe As traditional software providers, we face pressure. He said Adobe, long the standard for video and design software, has seen its share price slump in recent months, suggesting the company’s “all-in-one creative suite is vulnerable without all these creative AI marketing tools bundled together.”

PixVerse estimated annual recurring revenue at $40 million in October.

That’s not the only thing that generates revenue.

Kling, an AI video tool developed by TikTok’s rival Kuaishoposted nearly $100 million in revenue in the first three quarters of 2025, based on CNBC calculations from public information.

For now, Xie insisted that PixVerse is prioritizing technology development over commercialization and that the company has enough funding to operate for 10 years.

Addressing concerns about low-quality AI-generated content, often referred to as “slop,” Xie likened the current stage of development to the early stages of computer graphics, arguing that quality will improve as the technology matures.

“Initially there will be good and bad[content]but gradually the fittest will definitely survive…and some people will refine the technology to truly meet the human need for emotional and spiritual value.”