PixVerse raises $439 million in an extended Series C round

Video generation just minted another multi-billion-dollar claim. Singapore-based PixVerse told TechCrunch it closed a Series C extension that brings the full Series C to $439 million and pushes valuation past $2 billion. The cash is aimed at world-model work and a wider geographic push โ not another incremental clip demo.
That is a lot of money for a company founded in 2023. The market is treating high-quality video gen, and the leap into interactive worlds, as scarce infrastructure.
What the Series C extension actually closed
PixVerseโs own announcement says the extension completes Series C fundraising at USD 439 million. TechCrunchโs account matches the headline number and adds the valuation detail the company shared with reporters. The initial Series C closed in March, led by CDH Investments; Bloomberg had pegged that first tranche around $300 million, though PixVerse did not disclose the split publicly.
The company says it has more than 150 million registered users and over 15 million monthly actives across 177+ countries. It offers image-to-video at $4.80 per minute of generation, per TechCrunch, and declined to break out paid conversion. Headcount sits around 150 across Singapore, Beijing, and Shanghai, with hiring planned for research and go-to-market.
Who is on the cap table now
New names in the extension, per both PixVerse and TechCrunch: Alibaba, Lollapalooza Capital, Ivy Capital, Grand Mount Capital, Eastern Bell Capital, Mirae Asset, BlueFocus, and CloudAlpha. Returning: iGlobe Partners and OCBCโs Lion X Ventures (styled LionX in the company post).
Alibaba is not only a check. PixVerse already has a deal with the investor to deploy video-generation features. That is distribution, not just branding. Co-founders are Wang Changhu (CEO; former ByteDance computer vision) and Jaden Xie (president; formerly Lighthouse Capital).
V-Series and world-model roadmap stakes
PixVerseโs stack, as described to TechCrunch: a V-Series for consumer and API video, a C-Series for professional film and commercial work, and an R-Series of world models for games and world-building. The company plans a new V-Series model and a new world-model version this year. Flagship video model V6 is the current consumer/commercial face; R1, launched January 2026, is pitched as a real-time world model that turns fixed clips into continuous, interactive streams.
The company post pushes harder on games and live entertainment: a PixVerse Game Engine built on R1, shared worlds, personalized avatars, and real-time interactive livestreaming where AI characters respond to viewers. Xie told TechCrunch the edge is labeling discipline inherited from ByteDance-era visual understanding, not raw data volume. Take the claim as strategy, not audited proof.
Why video-gen valuations keep climbing
Capital is clustering because the shortlist of credible video labs still looks thin. Xie argued OpenAIโs Sora 2 exit and uneven quality from bigger platforms leave room for specialists. Competitors named in the coverage include ByteDanceโs Seedance, Kling, Video Rebirth, Midjourney, Runway, and Luma, plus world-model efforts from other high-profile labs.
Analysis: a $2B-plus mark on a three-year-old video startup is a bet that interactive generation โ not just prettier 4K clips โ is the next platform layer. Alibaba on the cap table and in the product pipeline is the sharper signal. If the new V-Series and R1 follow-through ship, PixVerse is buying time and distribution in a race that burns compute as fast as it burns cash. If they slip, this round becomes an expensive reminder that user counts are not the same thing as a durable model moat.



