"Zhengqi Future" Closes Hundreds of Millions of RMB in Angel Round Financing, Linear Capital Invests Across Both Rounds | Linear Portfolio

Today, physical AI company Zhenqi Future announced it has completed a multi-hundred-million-yuan angel funding series within eight months, with core investors including CDH Investments' VGC platform, Linear Capital and other top-tier financial institutions, as well as first-tier automotive industry capital such as SAIC's Hengxu Capital, Bairui Capital, and Zhengxuan Investment, plus premium consumer industry capital including Chow Tai Fook Investment.

The funds will be directed toward four priorities: R&D of the QUORRA DoorMind world model and QUORRA FLAT motion execution platform; mass manufacturing of QUORRA short-distance mobility robot models; building online channels including JD.com and Tmall flagship stores plus overseas DTC operations, alongside offline networks across more than ten countries globally and multiple provinces in China; and refinement of next-generation product designs.

Neil Zeng, Partner at Linear Capital, commented: "Zhenqi Future sits at a rare intersection: the cost feasibility enabled by China's smart driving supply chain spillover, the 'data oil field' of real living spaces in door-to-door scenarios that remains largely untapped, and a team that has 'say it, do it' carved into every milestone. We believe that as the flywheel of product, data, and model begins to spin, the company's moat will only deepen."

Zhenqi Future is a physical AI company focused on Door-to-Door (D2D) mobile task scenarios, with model capabilities as its core technology, entering real-world environments through essential products and building a data flywheel. The company pursues a dual "technology + product" driver strategy, with a path notably similar to Tesla's — using end products as entry points, continuously iterating models with real-world data, then feeding stronger models back into product experience to create a self-reinforcing growth loop.

Zhenqi Future believes its endgame is to become the underlying infrastructure for next-generation physical AI — a spatial intelligence foundation that all living-space mobile robots will have to navigate through.

Currently, Zhenqi Future is considered the only physical AI company actively capturing short-distance mobility data within unstructured indoor living environments, building technological and data barriers through this first-mover advantage. Its founder, Shulin Wu, previously served as VP of Baidu IDG's intelligent driving business and as General Manager at Huawei, bringing full-stack commercialization capabilities.

This closed loop is driven by the co-evolution of "brain" and "body."

Zhenqi Future's self-developed "brain," QUORRA DoorMind, is a rare and currently the only native foundation world model in the industry with continuous indoor-outdoor full-scenario perception capabilities, designed for universal mobility spaces. It possesses multimodal spatial autonomous cognition and intelligent evolution capabilities in D2D scenarios, breakthrough-solving the industry's long-standing pain point of fragmented indoor and outdoor scene perception.

The model deeply integrates multidimensional modal information including visual perception, natural language understanding, and environmental situational awareness. Through core capabilities such as real-time 3D spatial deconstruction, precise semantic feature parsing, and dynamic environment adaptive modeling, it achieves fine-grained understanding of 3D spatial scenes across the full D2D process with seamless indoor-outdoor continuity.

Based on complete, uninterrupted spatial semantic cognition of the real physical world, the model can build a comprehensive full-scenario semantic planning system, truly implementing integrated indoor-outdoor navigation without high-definition maps. For complex environmental interaction scenarios, it can execute precise intelligent decision-making and dynamic game-theoretic reasoning, achieving cross-domain collaborative adaptive operation across multiple scenarios and temporal contexts, creating full-stack, self-iterating, closed-loop-controllable D2D universal mobility autonomous decision-making capabilities.

QUORRA DoorMind is the industry's first native world model oriented toward Door-to-Door mobile tasks, achieving for the first time continuous perception and unified understanding of indoor and outdoor scenarios, breaking the industry's long-standing perceptual fragmentation. It goes beyond merely detecting obstacles, fusing visual, linguistic, and environmental information to truly understand spatial semantics and rules — this is a corridor, proceed straight; this is an elevator lobby, wait; this is a mixed pedestrian-vehicle zone, negotiate — thereby achieving integrated indoor-outdoor navigation without HD maps, and continuously evolving with every task.

If QUORRA DoorMind is the "brain," QUORRA FLAT is the "body" that translates commands into physical action. It features a full-domain by-wire chassis with four-wheel independent drive and steering plus adaptive active suspension, combined with RLHF technology for autonomous evolution and intelligent regulation of chassis dynamics. Leveraging deep learning of physical operating laws between vehicle body and road surface from the world model, it real-time deduces vehicle motion states, anticipates road dynamic disturbances, and iteratively optimizes multi-wheel collaborative control strategies to complete closed-loop iterative optimization of full real-vehicle operation.

It can comprehensively coordinate and schedule four-wheel torque output, independent steering angles, and suspension vertical support states, synchronously tuning body posture through three-axis coordination, breaking through the inherent limitations of traditional chassis fixed-calibration modes. It can adapt to complex conditions including urban roads, off-road scenarios, and low-adhesion surfaces, achieving full-domain dynamic optimization of handling, comfort, and safety.

The two are not in a simple "command-execution" relationship: QUORRA FLAT continuously learns from physical feedback in every movement and feeds it back to QUORRA DoorMind, jointly optimizing the complete loop from perception to execution. This collaborative architecture gives Zhenqi Future the capability to create truly high-value products that solve user pain points — the model lets the robot "see and understand," the body lets that understanding "land steadily." This is precisely the technological moat that distinguishes the company from pure algorithm companies or hardware assemblers.

Currently, most robot data remains concentrated in laboratories, enclosed facilities, or specific task scenarios, struggling to support model generalization in complex environments. QUORRA, by entering real user living scenarios, continuously captures a category of long-neglected "data oil fields" — indoor-outdoor continuous mobility data encompassing environmental changes, human-space interactions, and task feedback.

This "data oil field" is in fact one of the hardest nuts to crack in the physical world. The complexity of scenarios within the last few kilometers of living service radius near one's doorstep far exceeds imagination: needing to smoothly traverse cobblestone surfaces, pass through narrow doors, and in mixed pedestrian-vehicle, indoor-outdoor switching scenarios, also facing elevator signal interruption, road surface water accumulation, and sudden appearances of small dogs, delivery vehicles, or children — each obstacle corresponds to independent engineering challenges.

In traditional autonomous driving's SAE L1-L5 classification, system autonomy improves while the environment remains fixed as motor vehicle lanes. This grading struggles to cover the proposition that Door-to-Door mobile tasks must answer: "Can a robot depart from a specific doorway and autonomously complete the full process of cross-domain space, cross-traffic medium, cross-semantic task arrival" — indoor-outdoor switching, elevator interaction, floor positioning, non-road traversal, last-100-meter semantic navigation — these core complexities entirely exceed the descriptive scope of traditional grading.

Therefore, Zhenqi Future proposes a dual-domain classification framework: a two-dimensional matrix of "Autonomy Level (A-axis) × Scenario Domain Complexity Level (D-axis)" to redefine Door-to-Door mobility capabilities. Scenario domains escalate from structured indoor domains (D1) to open urban D2D domains (D5), covering unstructured living spaces that traditional autonomous driving has never addressed. The company currently starts from D3-A3, evolving toward D4-A4, with a long-term target of D5-A4/D5-A5 — achieving high to full autonomy in the most complex universal scenarios.

At this point, Zhenqi Future's commercial path becomes clear: from the outset, the company has pursued a dual "model + product" wheel drive — with QUORRA DoorMind world model as the core technological barrier, entering real user scenarios through short-distance mobility as a high-frequency essential product, with products continuously generating high-quality data during D2D task execution, data flowing back to train models, and stronger models driving product experience upgrades — a D2D-native "product → data → model → product" growth flywheel has already taken shape. Unlike the single data dimension of road scenarios, QUORRA captures cross-domain physical data from continuous indoor-outdoor spaces. As capabilities continue to generalize, Zhenqi Future will persistently create differentiated products through model capabilities, obtain exclusive data through product deployment, and build model barriers through data training. The three interlock and spin ever faster.

It is understood that QUORRA DoorMind world model and QUORRA FLAT motion execution platform have already entered testing phases across multiple scenario types, and with the formal commercialization of the first robot product, the data flywheel has already been activated.

Though still at the angel stage, Zhenqi Future's global commercialization path is already unfolding.

The debut product QUORRA X5 "launches with mass production, mass produces with global export." The product has already secured commercial orders from more than ten overseas countries and regions, with plans to develop overseas DTC business, while establishing Tmall and JD.com flagship stores and domestic multi-province offline sales networks within the year, directly targeting global consumer markets.

Relevant industry research reports indicate that current global short-distance mobility vehicle ownership exceeds 2 billion units, compounded by three growth drivers — AI intelligentization upgrades, cross-category substitution of passenger vehicles, and accelerated adoption in Asia-Africa-Latin America markets — with industry compound annual growth rate expected to reach over 26%, and global market size potentially exceeding $2 trillion by 2035. This market space provides vast opportunity for Zhenqi Future. These mass-produced products will function like data exploration vessels, sailing into real living spaces across the globe — what they constitute is not merely a sales network, but a continuously producing data oil field.

Despite the rapid commercialization development, Zhenqi Future maintains that its endgame is not to become "a hardware company that sells well," nor "a software company with impressive models." Its ambition is: with advanced models as the core technology engine, essential products as data entry points and commercial carriers, to build a self-reinforcing, continuously expanding "product → data → model" flywheel in global real living spaces.

"When this flywheel accumulates sufficiently deep data and sufficiently generalized models, Zhenqi Future will become one of the most defensible companies in physical world mobile intelligence — not because of any single product or model, but because of the flywheel formed by the interlocking of product, data, and model, where every additional rotation deepens the moat — this is a dynamic, self-reinforcing moat over time, not a static barrier vulnerable to single-point breakthroughs. And this, precisely, is Zhenqi Future's ultimate way of forging time into its moat," said founder Shulin Wu.