Back to feed
News Story
APriority70
TechCrunch AI
3 sources

Video generation startup PixVerse raises $439M, valuation soars past $2B

Original

新加坡视频生成初创公司PixVerse完成C轮扩展融资,融资金额4.39亿美元,估值超过20亿美元。该公司拥有1500万月活跃用户,此次融资彰显了投资者对AI视频生成领域的信心。

SynthePulse Insight · AI deep reading

PixVerse Valued at $2B: The 'Survivor' Logic in the AI Video Race

Version 2 · 3 sources

After OpenAI shut down Sora 2 and Meta and Tencent stumbled, Singapore-based AI video startup PixVerse has raised $439 million in Series C extension funding, pushing its valuation to $2 billion. This reveals investors still believe the赛道 can accommodate multiple winners, but the barrier lies in data labeling rather than data itself.

  • PixVerse completed a $439 million Series C extension round at a valuation exceeding $2 billion, with investors including Alibaba and Mirae Asset.
  • The company claims 150 million registered users and 15 million monthly active users, supporting video generation up to 4K resolution.
  • Co-founder Wang Changhu previously led visual AI technology at ByteDance's TikTok; his data labeling expertise is considered a core advantage.
  • OpenAI shut down Sora 2, and Meta and Tencent failed to deliver high-quality video models, creating a 'survivor wins' dynamic.
  • PixVerse plans to use the funds for global enterprise partnerships, new product launches, and team expansion (currently 150 employees).
Open section navigation1. Funding Data: A Signal of Counter-Cyclical Growth

1. Funding Data: A Signal of Counter-Cyclical Growth

In July 2026, Singapore-based AI video generation company PixVerse announced the closing of a Series C extension round totaling $439 million, with a valuation exceeding $2 billion. The round included new investors such as Alibaba, Lollapalooza Capital, and Mirae Asset, with existing shareholders iGlobe Partners and OCBC's Lion X Ventures continuing to follow. The earlier Series C round was led by CDH Investments in March 2026, reportedly worth about $300 million according to Bloomberg.

In the AI video generation field, this valuation is second only to the industry's leading players. Notably, OpenAI shut down its video generation model Sora 2 in early 2026, while Meta and Tencent were said by PixVerse co-founder Xie Jia to have 'failed to meet quality standards.' PixVerse's counter-cyclical funding shows that capital hasn't cooled despite the exit of giants; rather, it believes the market needs more specialized players.

2. Users and Products: Scaled but Not Profitable

PixVerse claims its consumer product has over 150 million registered users and 15 million monthly active users, but it has not disclosed the proportion of paying users. Its video generation service supports up to 4K resolution with audio, with a base price of $4.80 per minute for image-to-video generation. The company offers three product lines: the V-series model for consumers and API, the C-series model for professional film and TV production, and the R-series world model for game development, released in early 2026.

However, there is still a gap between user scale and commercial monetization. Competitors like Runway and Pika also have large user bases but unclear profitability models. PixVerse's advantage lies in its co-founder Wang Changhu, who previously led visual AI technology at ByteDance for TikTok. The company claims its fine-grained data labeling brings model quality advantages. Xie Jia said, 'The key is not data, but how you label it—data is everywhere.' This claim has yet to be independently verified.

3. Competitive Landscape: A Window of Opportunity After the Shakeout

The AI video generation track is undergoing a shakeout. Besides OpenAI's exit, Meta and Tencent are said to have encountered obstacles in high-quality video modeling. However, competition remains fierce: ByteDance has its Seedance model, former Tencent AI head Liu Wei founded Video Rebirth, and Asia also has Kling AI; in the West, Midjourney, Runway, and Luma continue to iterate. Meanwhile, startups from Yann LeCun and Fei-Fei Li are also building world models, directly competing with PixVerse's R series.

PixVerse's differentiation lies in its data labeling technology derived from core experience in short-video recommendation systems. But can this advantage be sustained? ByteDance itself is also pushing video generation and has stronger computing power and data pipelines. Moreover, PixVerse currently has only 150 employees (distributed across Singapore, Beijing, and Shanghai), which is small compared to the human resources of giants.

4. Future Path: Enterprise Partnerships and Organic Growth

PixVerse plans to use the new funds to expand global enterprise partnerships, and has already reached a video generation deployment agreement with investor Alibaba. In terms of products, the company plans to release the next-generation V-series video generation model and a new version of the world model within the year. For the team, it will focus on hiring researchers and market execution staff.

However, the enterprise market also faces challenges: compliance, customization needs, and price wars from competitors. Additionally, is PixVerse's $2 billion valuation reasonable? With 15 million MAUs, the per-user valuation is about $133; if considering paid conversion rates, the figure could be higher. Investors are clearly betting on the explosive application of video generation in marketing, education, and media, but a profitability timeline has not been specified.

Credibility boundary

The facts in this article are primarily sourced from reports by TechCrunch, The Decoder, and The AI Insider. The three sources mutually confirm core data (funding amount, valuation, user numbers). Comments from co-founder Xie Jia come from an exclusive TechCrunch interview and are directly quoted from the source reporting. Competitor analysis is based on public information. User numbers and paid data have not been independently verified.

Insight takeaway

PixVerse has gained significant capital recognition in AI video generation but faces dual pressures from giant encirclement and commercialization validation. Its data labeling advantage provides a short-term barrier, but long-term success depends on product iteration, market expansion, and cost control. Investors believe there is still room in this track, but 'survivors' still need to prove they are not 'survivorship bias.'