"UNICUS Closes New Funding Round Led by Linear Capital, Building Lego for the AI Era | Linear Portfolio"

Today, personalized manufacturing brand UNICUS announced a new funding round led by Linear Capital, with a leading founder from the Maker space participating as an individual investor for the first time.

Through sustained market education and continuous technical refinement, UNICUS has deeply integrated AI generation with flexible manufacturing, enabling users to upload a single photo and receive a bespoke physical artifact — capturing life's meaningful moments.

Zeren Bai, Senior Director at Linear Capital, said: "Linear has long paid close attention to real-world deployment opportunities for Generative AI. As AI makes personalized design and manufacturing scalable, this represents a massive opportunity to create 'the LEGO of the AI era.' We look forward to UNICUS continuing to define new consumer product paradigms driven by AI."

On April 13, personalized manufacturing brand UNICUS (formerly Fangzai Photo Studio, Shenzhen Qianzhi Technology Co.) completed a new funding round of several million USD. This round was led by Linear Capital, with Jiukun Venture Capital and FutureX Capital participating. The founding team and a leading founder from the Maker space invested as individuals.

The proceeds will primarily fund training of the brick generation foundation model, R&D of AI Agents, and overseas market expansion — continuing to build out the core competency of "AI-driven personalized manufacturing."

Over the past two years, generative AI's explosion has largely remained in the digital realm. UNICUS chose to bring AI into physical factories. Based on its self-developed fully automated design and production system, UNICUS pioneered a technology-driven path to mass customization of toys.

For decades, the global brick market has been dominated by giants like LEGO, with annual revenue of roughly 70 billion yuan. Yet traditional bricks are fundamentally "IP-driven standardized industry" — whether Disney castles or Star Wars sets, consumers purchase cultural symbols within predetermined frameworks.

In the view of UNICUS founder and CEO Hao Xu, the highlight moments, unforgettable memories, and precious time spent with family and pets in each person's life are the most authentic and unique "personal IP."

However, the customization industry faces a paradox: personalization is inherently "anti-scale." Traditional customization relies heavily on manual labor. Before large models emerged, after users uploaded photos, designers often needed one to two days of manual adjustment and asset library matching, with final results frequently limited by individual designer experience. High time costs and uncertain outcomes made customized products difficult to bring to mass consumers.

Over the past five years, through sustained market education and continuous technical refinement, UNICUS has deeply integrated AI generation with flexible manufacturing, enabling users to upload a single photo and receive a bespoke physical artifact — capturing life's meaningful moments.

In mid-2024, UNICUS's self-developed brick foundation model LEGO Maker officially entered production. This is the first large model to use autoregressive networks for general LEGO brick model generation, with related research published in Transactions on Graphics, the world's premier computer graphics journal. The model encodes each brick assembly as tokens for sequential prediction, supporting thousands of brick part types to generate structurally stable, physically manufacturable and assembleable brick models directly from images and text — achieving scalable customization of personalized products.

With the LEGO Maker model, UNICUS's products will no longer be limited to the single mode of "photo-based custom brick generation." Instead, they will be able to engage in dialogue with users and process image inputs, understanding the stories and characteristics behind the people to generate richer, more personalized scene-based effects.

For example, a user might input: "She loves skiing, she's a journalist..." The AI will combine image characteristics with the language description to automatically plan a 3D brick design featuring a skiing scene and profession-specific accessories. This evolves UNICUS's product from single figurines to more scene-rich, narrative "3D portrait sculptures."

On the production backend, UNICUS chose to build its own flexible factory. The challenge of customized manufacturing lies not in injection molding, but in sorting and printing. UNICUS maintains roughly 1,700 standard brick part types; how to quickly and accurately assemble hundreds of components from these parts for each unique order is a widely recognized industrial challenge. On the printing side, the real difficulty is: brick surfaces vary in shape, curvature, and arrangement, yet the production line must efficiently complete personalized printing that differs for every single order. This places high demands on printing scheme generation, equipment control, and production rhythm.

In its early days, UNICUS went through a phase of manual part picking, which proved extremely inefficient. The team then solved this pain point through a self-developed flexible sorting system and vibratory bowl feeder. The system automatically dispenses required parts from thousands of channels based on AI-generated instructions. Additionally, for the thousands of unique designs, UNICUS developed automated printing and machine vision quality inspection systems, achieving delivery capacity of nearly 10,000 distinct products daily with shipping as fast as 18 hours — roughly 10x overall efficiency improvement over traditional models.

Following this round, UNICUS will continue to dig deeper into "AI for CAM" (Computer-Aided Manufacturing), having AI directly generate manufacturing schemes to further solidify its moat in personalized manufacturing.

On the team side, founder and CEO Hao Xu holds a PhD in computer graphics from the Chinese University of Hong Kong, with over a decade in 3D generative design algorithms, multiple first-author papers at top international graphics conferences including SIGGRAPH, and 50+ domestic and international patents accumulated by the team. Co-founder Cheng Li holds a PhD in automation from Hong Kong University of Science and Technology, where he studied under Professor Zexiang Li, and previously led DJI drone intelligent production line R&D, responsible for hardware equipment and flexible manufacturing system construction.

Linear Capital has long paid close attention to real-world deployment opportunities for Generative AI. UNICUS uses AI to reconstruct personalized design and manufacturing capabilities, precisely targeting customized consumption scenarios. After keenly capturing market demand, the team became the first to close the full-stack loop from generation algorithms to flexible factory, achieving not only integrated design-production but also early commercial validation. We believe that as AI makes personalized design and manufacturing scalable, this represents a massive opportunity to create "the LEGO of the AI era." We look forward to UNICUS continuing to define new consumer product paradigms driven by AI.