A 29-Year-Old Gamer's Company Just Became China's Newest AI Unicorn — Vast Eyes Text→3D With $200M

A company that spits out 3D models from one line of text just became a unicorn

On June 1, Bloomberg reported that Beijing-based 3D-generation AI startup Vast raised nearly $200M to become "China's latest AI unicorn" at a $1B+ valuation. The founder is 29, a former gamer. A company that started from the question "why is making 3D assets so hard?" while gaming reached unicorn ranks in just a few years.

Here's why it's interesting: the spotlight in generative AI has been on text (LLMs) and images/video. 3D — needed across games, film, the metaverse, and manufacturing — remained an area that was simply "too hard to make." Producing a single 3D model could take a specialist artist days. Vast's flagship product, Tripo, breaks that wall — feed it text or an image as a prompt, and out comes a 3D model. Capital piled into that "getting easier."

Meet the players — Vast, Tripo, and founder Song

Vast was founded in 2023. Founder Song started the company alongside technologists from SenseTime, one of China's leading computer-vision/AI firms. So talent that emerged from there jumped into the nascent field of 3D generation. The "29-year-old former-gamer founder" story gets the headlines, but behind it sits solid vision-AI engineering talent.

Tripo (Studio) is Vast's flagship. Feed it text ("a knight in medieval armor") or an image, and it generates a matching 3D model. It's used by game developers, 3D artists, product designers, and even general users who just "need some 3D thing." It reportedly has about 10M individual users plus ~90,000 studios and companies. That means it's already a practical tool, not an "experimental toy."

The investor list is glittering, too. This round was co-led by Ince Capital and a China Life Insurance–backed venture fund, with Genesis Capital and existing backers Eminence and Primavera. On top of that, Chinese Big Tech Alibaba and Baidu are listed as investors. Big Tech betting directly is a signal that 3D generation is strategically important, not a mere niche.

The details — numbers and real-world deployment

Here are the key figures in a table.

Item	Value	Note
Raise	~$200M	Per June 1 Bloomberg report
Valuation	$1B+	Unicorn status
Individual users	~10M	Tripo
Company/studio customers	~90,000	Tripo
Co-leads	Ince Capital + China Life–backed fund	Genesis, Eminence, Primavera participating
Strategic investors	Alibaba, Baidu	Chinese Big Tech
Founded	2023	With ex-SenseTime talent

More impressive than the numbers is the real-world deployment. Vast supplies its tech to NetEase's RPG "Where Winds Meet." A player uploads their photo, and an interactive 3D avatar is generated on the spot. So "text→3D" runs not as a lab demo but inside a game enjoyed by millions. That's Tripo's real edge — engineering that runs "in real time, at scale," not just quality.

3D generation is hard because it has one more dimension than images. A 2D image is a plane of pixels, but 3D requires geometry, texture, and a clean mesh usable in game/rendering engines to all line up. Tripo gathering 10M users means it has, to some degree, crossed this "practicality" wall — which is why capital assigned it a unicorn valuation.

Who gains what — Vast, the game/manufacturing industries, China's AI ecosystem

For Vast, $200M buys both "GPU armament" and "land-grab." 3D-generation models cost enormous compute to train, and raising quality means continually pouring in data and computation. This funding accelerates model improvement and global expansion, and gaining Big Tech allies in Alibaba and Baidu secures favorable ground in cloud and distribution too.

For the game, manufacturing, and e-commerce industries, the direct benefit is "the collapse of 3D-asset production costs." Game studios can crank out characters, props, and environment assets fast; e-commerce can produce 3D product views, manufacturing can produce prototype geometry — in an instant. When work that took specialist 3D artists days shrinks to minutes, small teams that had given up on 3D over cost can enter the market.

For China's AI ecosystem, it's a narrative of "while the U.S. leads on LLMs and commercialization, we mint unicorns in applied areas like generative media and 3D." A few days ago Hacker News debated heatedly whether "the real measure of the AI race is commercialization or applications" — and cases like Vast show China surging ahead in applied/consumer products. Alibaba and Baidu investing directly carries that strategic intent.

Historical parallels — what happened to generative-media startups?

There have been many attempts to automate media production with generative AI. Their trajectory helps gauge Vast's future.

Success — the explosion of image/video generation. Companies like Midjourney and Runway grew fast by demolishing creative barriers with "text→image/video." The key was that "the moment quality crosses a threshold, users explode." Lesson: generative media adoption jumps nonlinearly once quality passes "usable" — and Tripo's 10M users signal that 3D is approaching that threshold.

Cautionary — the "great demo, useless in production" trap. Conversely, early 3D-generation tools had flashy demo reels but, plugged into a real game engine, often produced broken meshes or messy textures that pros couldn't use. Lesson: in 3D, "pretty results" and "results that actually enter the pipeline" differ. Vast actually shipping into a NetEase game is evidence it cleared this trap to a degree — but whether that holds across all use cases needs more validation.

Comparison — absorbed by Big Tech or swept up in price wars. Plenty of high-flying generative-media startups were eventually outpaced or acquired by Big Tech's in-house models. It's hard for an independent startup to hold out against giant platforms offering similar features at near-free prices. Lesson: Alibaba and Baidu being both investors and potential competitors is double-edged. If Vast can't build "irreplaceable 3D quality and ecosystem," its Big Tech allies could flip to competitors at some point.

Competitor counter-plays

U.S./Western 3D-generation startups counter with "integration into global game and Hollywood pipelines." They emphasize seamless hookups with mainstream game engines like Unreal and Unity, and with film VFX workflows. If Vast is strong in Chinese domestic gaming (NetEase), Western rivals will try to split the market by penetrating "global AAA game/studio standards."

Big Tech's in-house 3D models are the biggest threat. Google, Meta, and NVIDIA all invest in 3D/world generation, and NVIDIA in particular holds both the hardware and software sides of 3D content/simulation. If they bundle 3D generation as free/included, independent startups' pricing power weakens. Vast must differentiate with "overwhelming expertise in specific use cases (game avatars, real-time generation)."

Traditional 3D-asset marketplaces and artists are a different kind of pushback. They stress that "AI-generated 3D is weaker on copyright, quality, and consistency," arguing the premium market will stay with human artists. Indeed, high-quality, original 3D needs human hands for now, and AI is more likely to take the "high-volume, low-cost, fast" segment first.

So what actually changes

For game and app developers, it signals "the era of outsourcing or hand-modeling 3D assets is ending." If tools like Tripo can crank out characters and props in minutes, even small teams can make 3D-rich games and apps. That said, you have to validate per-use-case whether "AI-generated meshes run cleanly in your actual engine," and quality-critical parts will still need human touch-up.

For founders and investors, the key is that "unicorns emerge in applied generative AI, not just LLMs." The text-model race has become a Big Tech capital war, but in specific modalities like 3D, voice, and avatars, specialized startups can build markets fast. In China, a pattern of Big Tech (Alibaba, Baidu) investing directly in such startups has taken hold, and it's becoming an engine that mints "applied-AI unicorns" quickly.

For readers watching the industry, the big picture is that "the AI competition's front is widening beyond text into 3D, space, and the physical world." It started with in-game avatars, but the same tech will spread into the metaverse, robot simulation, digital twins, and manufacturing design. Note that the "~$200M, unicorn" figures are per a Bloomberg report on a private round, so exact valuation and terms are best confirmed by a later official announcement.

References

• Bloomberg — Gen-Z gamer's 3D-model startup becomes China's latest AI unicorn
• Implicator.ai — Vast raises nearly $200M for Tripo AI 3D models
• Business Standard — Gen-Z gamer's 3D startup raises $200M, becomes China's AI unicorn