Math Magic Raises Nearly $50M Across Two AI 3D Rounds

Math Magic has closed a Series A+ and says it raised nearly $50M across two rounds in six months, giving the Beijing company more capital to connect AI-generated 3D models with physical production and global fulfillment.

BAI Capital and HyT Capital led both rounds. V Fund participated alongside existing investors HSG, IDG Capital, Long-Z Investments, Crystal Stream, and Jinqiu Capital. Math Magic did not disclose the individual size of the Series A+ or separate the nearly $50M total between the two financings.

The company is building across software, commerce, and manufacturing. That is a larger opportunity than an AI modeling tool, but it also leaves Math Magic responsible for every handoff between a digital object and the product that reaches a customer.

What Math Magic Announced

Math Magic announced the Series A+ close on August 20, 2026. The company said its aggregate financing across two rounds completed within six months was nearly $50M.

Both rounds were led by BAI Capital and HyT Capital. V Fund joined the current financing, while HSG, IDG Capital, Long-Z Investments, Crystal Stream, and Jinqiu Capital were identified as existing investors.

The accounting boundary is important. Nearly $50M is a cumulative amount across two rounds, not the disclosed size of one Series A+ round. The announcement does not identify a valuation, ownership terms, the exact prior-round amount, or the exact Series A+ amount.

The Product Stack Behind the Financing

Math Magic operates three connected product layers. Hi3D is an AI 3D creation platform. Builda is a creative community and commerce layer for personalized physical products. Builda Camera lets a user upload a photograph, generate a personalized figurine, and order the finished product.

Hi3D includes image-to-3D generation, texturing, model splitting, multiformat export, and tools designed for 3D printing. The company says the latest model improves geometric fidelity and reduces the early modeling and repair work required from artists.

Those capabilities address the digital side of the workflow. Builda and Builda Camera extend the responsibility into customer acquisition, personalization, manufacturing, quality control, fulfillment, returns, and delivery.

That end-to-end structure differentiates Math Magic from companies that can stop when a usable file appears on a screen. It also creates a company with the operating complexity of a software platform, a creator marketplace, and a custom-manufacturing network at the same time.

Why Generation-to-Delivery Is the Real Thesis

Math Magic describes its model as generation-to-delivery. The phrase captures the commercial ambition: make a personalized physical product feel almost as immediate as generating an image.

The difficult work begins after the prompt. A 3D model may need to be split into printable parts, textured, translated into compatible production files, manufactured with consistent color and detail, inspected, packaged, and delivered across markets.

Personalization increases the challenge because each order can carry unique geometry, customer expectations, and production requirements. A standardized consumer product can optimize a repeatable line. A personalized object has to preserve uniqueness without destroying throughput or unit economics.

The company is therefore betting that model quality and supply-chain integration can compound. Better generation can reduce manual production work. Better manufacturing data can inform product design. A creator community can supply demand, while fulfillment capability can turn digital engagement into revenue.

How Math Magic Plans to Use the Capital

Math Magic said the new capital will support its core AI 3D technology, continued product development, and the manufacturing and fulfillment infrastructure required for global expansion.

Technology investment can improve model quality, speed, texturing, and production readiness. Product investment can deepen Hi3D, Builda, and Builda Camera. Manufacturing and fulfillment spending must make the physical side of the system reliable enough that customers experience one product rather than a chain of loosely connected vendors.

The capital will also support global expansion. That introduces additional work around logistics, local production partners, payments, consumer expectations, intellectual-property controls, and the economics of shipping highly personalized objects across borders.

The Investor Bet

BAI Capital and HyT Capital led both financings, giving the company continuity across a fast six-month capital cycle. The repeated lead positions suggest the investors are underwriting more than a model-performance benchmark.

The thesis is that AI can reorganize how physical products are conceived, produced, and sold. Instead of manufacturing beginning with a forecast and a standardized catalog, production can begin with individual demand generated through software.

That argument is attractive because it connects generative AI to real commerce. It is difficult because every advantage in personalization can create new costs in production, fulfillment, customer support, and quality assurance.

What Math Magic Still Has to Prove

Math Magic must show that model quality survives the transition into a physical object and that the supply chain can preserve speed, fidelity, and margin as order volume grows. It also has to demonstrate that Builda can become a durable community and commerce layer rather than a temporary novelty around AI-generated figurines.

The company has not disclosed revenue, order volume, unit economics, manufacturing capacity, or the geographic mix of current users. Those missing metrics limit any claim about commercial scale.

Nearly $50M across two rounds gives Math Magic meaningful capacity to build. It does not collapse software, manufacturing, and fulfillment into one solved problem.

The prompt is the beginning. The business starts when something has to arrive.