Singapore-based video-generation startup PixVerse just closed a Series C extension worth $439 million in total, pushing its valuation past $2 billion. The company confirmed the raise to TechCrunch AI, saying the fresh capital will fund its world model work and a broader push to reach customers across more countries. This is one of the biggest AI video funding rounds we’ve seen this year, and it lands at a moment when the market is separating the serious players from the hopefuls.
What happened
PixVerse first closed its Series C back in March, led by CDH Investments. The amount wasn’t disclosed, though Bloomberg pegged it around $300 million. The extension announced now adds a wave of new backers, according to TechCrunch AI:
- Alibaba, Lollapalooza Capital, Ivy Capital, Grand Mount Capital
- Eastern Bell Capital, Mirae Asset, BlueFocus, CloudAlpha
- Returning investors iGlobe Partners and OCBC’s Lion X Ventures
The company was founded in 2023 by Wang Changhu and Jaden Xie. Changhu worked on computer vision at ByteDance. Xie came from investment firm Lighthouse Capital. That ByteDance pedigree matters, and I’ll get to why in a second.
The product lineup
PixVerse runs three model families:
- V-Series for consumer and API video use
- C-Series for professional film and commercial workflows
- R-Series world models for game development and world building, released earlier this year
Users can generate video up to 4K resolution with audio baked in. The consumer product claims over 150 million registered users and more than 15 million monthly active users. PixVerse wouldn’t say how many pay, but it charges $4.80 per minute for image-to-video generation, which is competitive for the category.
Why this matters
Xie’s pitch isn’t that PixVerse makes “high-quality” video. Everyone says that. His argument is about labeling. “We think the key difference is not in data, but how you label it, because data is available everywhere,” he told TechCrunch AI. He credits his co-founder’s ByteDance experience building the visual understanding tech behind TikTok’s recommendation engine. The claim: accurate data labeling is the real moat, and that skill transfers directly to training video models.
What stands out here is the confidence about the competitive field. Xie argues only a handful of companies clear the quality bar. “OpenAI exited the business when they shut down Sora 2. Other companies like Meta and Tencent are not able to create high-quality video models,” he said. That’s a bold read, and worth watching, because the field is anything but empty.
The competitive picture
PixVerse is stepping into a crowded ring. From Asia, it faces ByteDance’s Seedance, former Tencent AI head Dr. Wei Liu’s Video Rebirth, and Kling AI. In the West, Midjourney, Runway, and Luma are all shipping. And the world model race is filling up fast, with startups from Yann LeCun and Fei-Fei Li building in the same direction.
So the money buys PixVerse room to move, but it doesn’t buy quiet. This is significant because it signals that investors still see video generation as an open contest, not a settled one, even after OpenAI pulled back Sora 2.
What comes next
Here’s what PixVerse says it’s planning with the new cash:
- Launch a new V-Series video model this year
- Ship an updated version of its world model
- Expand enterprise outreach globally, including a deal to deploy video features with investor Alibaba
- Hire more researchers and go-to-market staff on top of its current 150 employees across Singapore, Beijing, and Shanghai
For practitioners, the takeaways are practical. Enterprise buyers now have another well-funded vendor pitching creative, learning, and marketing use cases. Developers get another API option in a market where pricing and quality are moving fast. And the Alibaba deployment deal is a real distribution channel, not just a logo on a slide.
The video-generation market is heating up, and PixVerse just gave itself a war chest to compete. Whether labeling really is the moat Xie claims will show up in the next model release, not the press release. You can read the full details at the original TechCrunch AI report.