The Zen Gamer Building a World in 3D

Somewhere between a Zen temple in the hills near Hangzhou and a lecture hall at Johns Hopkins where the syllabus involved Middle East politics and the Arabic alphabet, Simon Song picked up a habit that most venture capitalists would find alarming in a founder: he thinks about first principles before he thinks about markets. It has not hurt him. At 28, Song runs Tripo AI - a company registered under the rather grand name VAST - and the phrase people keep attaching to it is "world leader in 3D AI." He would not dispute it. "For AI 3D large models, we are the leader in the world," he says. "I think there's consensus in that."

The consensus, if there is one, arrived quickly. Song founded VAST in 2023. Less than three years later the company's platform, Tripo Studio, has passed 6.5 million users, crossed $12 million in annual recurring revenue, and signed up a client list that reads like a directory of the entertainment-industrial complex: Tencent, NetEase, Sony, Microsoft, HTC, Stability AI, and - the detail everyone remembers - Pop Mart, the company behind the Labubu dolls currently colonising the world's tote bags.

OriginsThe boy who read the Tao Te Ching for fun

Song was born in 1997 in Hangzhou, and his childhood reading list was not typical. He found the Tao Te Ching and the Platform Sutra in elementary school, and by middle school he was practicing Zen at Jinshan Temple. It is tempting to treat this as colour, the sort of biographical detail that gets sprinkled into a founder profile to make an engineer seem interesting. But Song's philosophical bent turns out to be load-bearing. He talks about 3D not as a file format but as "a cornerstone of multimodal scenarios," and he frames his company's mission in terms borrowed from the early Chinese Marxist thinker Li Dazhao: to advance civilisation for the world, to create happiness for humanity. Grand words for a business that, on a practical level, turns the word "Superman" into a downloadable mesh.

He went to the United States for more than eight years, first at The Stony Brook School and then at Johns Hopkins, where his interests spanned religious philosophy, Hebrew, Arabic, and gaming, all pursued with what one profile called "equal intensity." The gaming part matters. Ask Song why he does what he does and he does not reach for a slide about total addressable market.

The pivotA Google demo, and a decision

Before VAST, Song had a career most founders would happily retire on. He worked in the CEO's office at SenseTime, one of China's original AI pioneers, running work on AI gaming and 3D animation. Then he co-founded MiniMax, the unicorn behind the Hailuo AI models and one of the country's celebrated "AI tigers." He was, by any reasonable measure, set.

What unsettled him was a Google research release. In 2022 the company published DreamFusion, a text-to-3D model that was rough and slow and mostly a curiosity to the wider world. Song saw a starting gun. Inside SenseTime he had watched top-tier creators burn absurd amounts of time on labour-intensive modelling work, and he had noticed a gap - a translation problem - between the people with ideas and the engineers who could realise them. DreamFusion suggested the gap could be closed by a machine. He left, gathered collaborators, and set up VAST in Beijing shortly after.

The productFrom toy to industrial tool

For a while, Song is refreshingly happy to admit, AI 3D was junk. "Before this year, AI 3D was more like a toy," he says. The turn came with Tripo 3.0, released in 2025 with what the company calls its most accurate, detail-rich geometry yet. Song's framing is characteristically blunt: "With Tripo 3.0 it has become a pipeline-ready industrial tool." Translated out of founder-speak, that means the models it produces are clean enough that a game studio or a 3D-printing shop can actually use them without a human spending hours fixing the topology afterwards.

The metrics that make investors lean in are less about the models and more about the business under them. VAST reports 40,000-plus studio and corporate partners, over three million professional users, and a research output - 50-plus papers, 30-plus at top AI conferences - that would flatter a well-funded lab. And then the line founders love to drop and rarely can: "It's a profitable business right now." In an industry where profitability is often treated as a rumour, Song says it like a fact.

He is unromantic about what all this technology is for. His view of quality is almost cruel in its clarity, and it is the closest thing he has to a religion. He does not want to make 3D easier for professionals. He wants to remove the professionals from the equation entirely, so that the person with taste but no software skills can make the thing in their head.

The betEveryone types words. Soon everyone makes 3D.

Song's stated ambition is to let everyone create 3D content with zero barriers and zero cost. The comparison he keeps returning to is the internet's most democratic moment. "It's like when everyone could type words and you got Twitter," he says. He believes the same unlock is coming for three-dimensional content: give ordinary creators a tool that handles the craft, and a new kind of platform appears on the other side. Inside the company, the shorthand is a "3D TikTok" - a place where user-generated 3D flows the way short videos do now.

Whether that future arrives on his timeline is an open question, and Song, to his credit, does not pretend otherwise. "It's not perfect now," he says of the technology. "That is the beginning of our journey." His pitch to sceptics is disarmingly simple, the confidence of a gamer who has read the patch notes: "Try our 3D generation technology and compare the results. Ours will be better."

TimelineThe road to 3D

There is a neat symmetry to Song's story, though he would probably deny building it on purpose. A kid who wanted to make his own RPG and could not, who studied ancient texts about liberation, who then spent his twenties building a machine designed to liberate creativity from technical skill. "What they need are ideas and a sense of beauty," he says of the creators he is chasing. "AI will do the rest for them." It is either the most idealistic thing a 28-year-old CEO has said this year, or the most commercially astute. With Song, it is usually both at once.