Meshy has raised nearly $400 million at a $1.5 billion valuation, and the useful question is no longer whether AI can make 3D objects. It is whether studios, printers, and product teams will trust those objects outside the demo.
Meshy just put a serious price on AI-generated 3D. The company said in a July 21 PRNewswire release that it raised nearly $400 million in a Series B round at a $1.5 billion valuation, the largest round to date for a company built specifically around AI 3D. That's a real line in the sand. If you build games, 3D-print products, or prototype physical goods, this is the funding round that tells you the category has moved from clever output to business infrastructure.
The basic promise is brutally simple: give Meshy a line of text or one image, and it can generate a usable 3D model in about a minute. The company says work that once took specialist skills and expensive software over weeks now takes about a minute and roughly a dollar. The claim is big. It also comes with numbers behind it: Meshy says annual recurring revenue is growing about 12x year over year, with more than 12 million registered users and over 100 million models created.
PRNewswire carried the company announcement, but the investor names came through trade and funding trackers. TechFundingNews reported that the round was led by IDG Capital, Matrix Partners China, and Monolith Management, with existing backers Granite Asia, HongShan, BAI Capital, and Source Code Capital oversubscribing their allocations. Dealroom lists the same July 2026 round at a $1.5 billion valuation. The investor list is even louder. This is mostly China-linked venture capital placing a giant bet on a Silicon Valley 3D generation company founded by MIT PhD Ethan Hu.
The real test is production, not prompts
Meshy says five of the world's ten largest technology companies by market capitalization or valuation have teams building with its software. Keep the wording tight there. The company did not name those five companies, and it did not say they were all paying customers in the way a finance chief would use that phrase. It did name customers and partners across more checkable markets, including Nexon, NetEase Games, 37 Interactive Entertainment, Bambu Lab, Creality, Elegoo, FlashForge, xTool, Hugo Boss, and Sweden's museum of art and design.
That's not a small operating footnote. Gaming studios burn through characters, props, environments, and early concepts faster than traditional art teams can comfortably supply them. 3D printing companies need clean geometry that can survive slicers and actual machines, not just look decent on a screen. Meshy is selling into both of those problems at once, which is why the dollar-per-model claim matters more than the usual AI excitement around speed.
The caveat is obvious. A usable model is not automatically a final asset. Ask any game artist who has cleaned up topology, UVs, textures, rigging, and collision meshes after a rushed export. Meshy knows this, which is why its new product details lean hard into production language rather than prompt magic. The company says its 3D Agent can export in formats including FBX, OBJ, GLB, and STL, while its 3D printing workflow reaches a slicer success rate of up to 97%.
Meshy wants the workflow, not just the model
Don't miss the timing. Meshy launched its 3D Agent Beta on June 4, then followed with the funding announcement less than seven weeks later. The agent is meant to turn a conversation, photo, sketch, or text prompt into a downloadable 3D model, while also brainstorming concepts in batches and answering 3D creation questions inside the same chat. Here's the thing. That is not just a nicer interface. It is Meshy trying to sit earlier in the creative workflow, before a user has already decided exactly what object they want.
The July release also introduced Auto Split, Smart Topology, and 8K Texture. Auto Split breaks generated models into printable parts and arranges them on the print bed. Smart Topology generates cleaner geometry in about 10 seconds, with polygon counts from 100 to 15,000. 8K Texture is rolling out soon. None of those features has the neat viral appeal of typing a prompt and watching an object appear. They matter because they attack the ugly work that decides whether an AI asset actually enters a pipeline.
That changes the budget conversation. If Meshy stays in the prototyping slot, it is useful but replaceable. If it becomes the place where an indie studio builds its asset sets or a product team moves from sketch to physical mockup, the $1.5 billion valuation begins to make sense. You do not need to believe every marketing claim to see the direction of travel.
The product story is less tidy. The company release is doing plenty of its own selling, and there are still unanswered questions around enterprise retention, IP risk, and how much clean-up professional teams still need after export. That is still early. But a nearly $400 million Series B is not priced for novelty. It is priced for a future in which 3D content production becomes cheap enough that teams make far more of it.
Also read: AI is finding software vulnerabilities faster than security teams can fix them • OpenAI picks Dublin as its EU headquarters with €105 million and 250 new jobs • Microsoft's AI security overhaul sets a benchmark every enterprise security startup now has to beat
