IMPLICATOR
“The real shift happens when AI-generated 3D assets require no reconstruction before entering production workflows,” Simon Song said in Tripo AI's March 9 release for its P1.0 model. On Monday, Bloomberg reported that Song's Beijing company, Vast, raised nearly $200 million and crossed a $1 billion valuation, citing a company statement and an interview with the founder.
Vast's valuation rests on a narrower test than most AI funding stories. The company has to show that generated 3D assets can survive the topology, UV, material, collider and version-control steps Song later described to 80 Level, instead of stopping at attractive demo objects.
PE Daily reported June 1 that Vast completed A+ and A++ rounds totaling nearly $200 million, led by Ince Capital and China Life's Yangtze River Delta tech fund. The financing followed Tripo's March announcement of a $50 million round backed by Alibaba and Baidu Ventures. PE Daily said the new money would go mainly to AI 3D and general world-model talent, core algorithm work, data accumulation, global expansion and industrial partnerships.
Key Takeaways
- Vast raised nearly $200 million and crossed a $1 billion valuation.
- Tripo says P1.0 can generate production-ready meshes in as little as two seconds.
- Project Eden is the next test for persistent, reusable AI-generated worlds.
- The risk is whether studios keep AI 3D assets inside production pipelines.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
What the financing disclosed
The company statement identified Ince Capital and a China Life-backed venture fund as lead investors, with Genesis Capital, Eminence Ventures and Primavera Venture Partners also participating. PE Daily added Shenzhen's AI terminal industry fund, linked to Honor, Shanghai Semiconductor Investment, Shenzhen Capital Group and several other Chinese investors. The list puts Vast's 3D work in front of insurers, local government-linked funds and semiconductor capital, not only venture firms.
Follow the next AI infrastructure bet
Strategic AI news from San Francisco. No hype, no "AI will change everything" throat clearing. Just what moved, who won, and why it matters. Daily at 6am PST.
No spam. Unsubscribe anytime.
Song gives those investors a founder profile they can explain quickly. Forbes said he studied economics and international studies at Johns Hopkins before returning to China in 2019, working in the CEO's office at SenseTime and co-founding MiniMax. The fundraising report noted that he spent entire days at Johns Hopkins playing Civilization before his interest in animation and games led him toward 3D models. Ince founding partner Steven Hu put the investor case in personal terms. “Simon is the type of CEO China's tech-driven startups desperately need,” Hu said.
Song gave the Monday report the largest user count in the file: 20 million users, with most in the U.S. and the next groups in Europe, Japan and South Korea. Earlier source counts were narrower. Tripo said in March that it served more than 6.5 million creators and 90,000 developers, with nearly 100 million 3D assets generated. Forbes put the May 27 base at nearly 10 million individual users and 90,000 studios and companies.
The revenue model is already visible, but still early. The company charges subscriptions starting at $20 plus token fees, while Forbes said individuals pay $20 to $140 a month and corporate customers are charged by project.
Where the mesh enters production
The March 25 release centered on P1.0. Tripo said the model can generate production-ready polygon meshes in as little as two seconds, compared with workflows that can take minutes or still require retopology. The same release said the model was trained on about 50 million high-quality 3D assets, which Tripo described as one of the largest structured polygon-mesh datasets in the field.
In the 80 Level interview, Song described what happens after generation. Tripo assets can be exported as FBX, OBJ and GLB files, then moved into Blender, Maya, Unity or Unreal Engine, he said. He also listed the ordinary production steps that remain: initial screening, DCC refinement, engine testing and version control. The tool still sits at the front of a longer production path, according to Song's description.
The caution is market-based. Commercially viable breakthroughs remain rare in a gaming industry worth about $200 billion and still skeptical of the technology, the report said. Song's 80 Level interview gives the production reason. Teams still optimize topology, control polygon counts, adjust UVs and materials, add colliders and LODs, then put assets into version control.
Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.
Project Eden and the world-model race
Project Eden could be released as soon as this month, according to the Monday report. It described the project as a world model for explorable 3D environments rather than flat video frames. PE Daily said Eden separates state simulation from visual rendering, with a persistent world state that can be reused and support multiple users.
The category is already crowded. The Implicator has covered the earlier funding race around world models. Fei-Fei Li's World Labs raised $230 million in 2024 and another $1 billion in February 2026 for spatial-intelligence work. Tencent had released HunyuanWorld 1.0 by the time of the SCMP interview; by April, HY-World 2.0 materials described an open world-model framework with released weights and code.
Song told SCMP last August that Vast led the AI 3D model field. “For AI 3D large models, we are the leader in the world. I think there's consensus in that,” he said, when Vast had more than 100 employees and about 40,000 enterprise API customers. The newer account describes a 150-person company, with NetEase using its technology in Where Winds Meet to let players upload photos and generate interactive 3D avatars.
The consumer plan remains on the calendar. Song told Forbes that “AI has released people's creative impulses,” and he has described an “interactive TikTok” where users play short generated experiences instead of scrolling video. Joy Dai of Eminence framed the investor case around that shift, saying “the internet's third revolution could be interactive content,” with AI-tailored material and 3D as a new medium.
If Vast publishes Eden this month, the details to check are whether developers can store scene state, reopen edited environments and support multiple users, the three behaviors PE Daily attributed to the project. The studio test is separate: whether teams keep Tripo assets in production after the first trial.
Frequently Asked Questions
What is Vast?
Vast is the Beijing company behind Tripo AI, a platform that generates 3D assets from text and image prompts for creators, developers and studios.
How much did Vast raise?
Vast raised nearly $200 million and crossed a valuation above $1 billion. PE Daily said the company completed A+ and A++ rounds totaling nearly $200 million.
What is Tripo P1.0?
Tripo P1.0 is the company’s production-focused 3D model. Tripo says it can generate polygon meshes in as little as two seconds, though that figure comes from a company release.
What is Project Eden?
Project Eden is Vast’s world-model project for explorable 3D environments. The company says it separates world state from visual rendering so scenes can persist and be reused.
Why does this matter for game studios?
Game studios need usable topology, UVs, materials, colliders and version control. Vast’s claim is that AI-generated assets can enter that pipeline with less repair work.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai
