Fourier's most recent disclosed round was a Series E+ corporate minority investment of CNY 300M (approximately $43M), closed August 1, 2025, led by Runyang Technology.

It followed a Series E in January 2025 with participation from Guoxin Investment, Pudong Venture Capital, Zhangjiang Science Investment, Zhangke Yaokun Fund, Prosperity7 Ventures, and Junshan Capital. Earlier rounds included a Series D in January 2022 led by SoftBank Vision Fund 2, with participation from Prosperity7 Ventures and Vision Plus Capital, and a prior round backed by Qianhai Fund of Funds.

Fourier has raised approximately $246M in total disclosed funding across its history.

Fourier sells two connected hardware lines: rehabilitation robots for hospitals and therapy clinics, and full-size humanoid robots for research institutions, developers, and early commercial deployments in care and industrial settings.

The rehabilitation business is the more commercially mature line. RehabHub connects 30+ device types, including upper-limb systems, lower-limb exoskeletons, ankle and wrist trainers, and balance platforms, into a single software environment. Clinics can deploy multiple Fourier devices through RehabHub and give therapists a unified view of patient progress across sessions, while the robots perform repetitive, measurable guided-movement training and therapists supervise and adjust treatment. MetaMotus Galileo extends the platform into biomechanics analysis, combining a 6-DOF motion base, force plates, wireless muscle sensors, and a 180-degree LED screen running mixed-reality therapy scenarios, closer to a motion science lab setup than a single rehab machine.

The humanoid line spans three generations in the GR series. GR-1 is a 165 cm, 55 kg humanoid with up to 44 degrees of freedom, built for reception, guidance, and light industrial tasks. GR-2 increases that to 53 DOF with a newer actuator generation and optional 12-DOF dexterous hands, targeting developers and researchers that want to train the robot, customize it through SDKs, and integrate it into embodied AI workflows. GR-3, unveiled in August 2025, is the newest product in the line, a care-oriented humanoid with 55 DOF, a four-microphone array, 31 pressure sensors for touch, structured-light vision, and a dual-path interaction system that handles simple commands through fast rule-based responses and routes richer conversations through an LLM layer. Its design targets eldercare, hospital, and public-service settings rather than industrial use cases.

N1 sits alongside the GR series as Fourier's open-source humanoid platform: 1.3 m, 38 kg, 23 DOF, and capable of running at up to 3.5 m/s. Fourier released N1 alongside the FourierActionNet open-source dataset in 2025. The developer documentation center includes teleoperation tools, reinforcement learning motion-control code, simulation assets for NVIDIA Isaac, URDF and MJCF robot models, and SDKs for the dexterous hand and actuators, letting a university lab or robotics startup simulate locomotion policies and deploy them on real hardware without building the full stack from scratch.

Across both product families, Fourier uses its self-developed Fourier Smart Actuators in the GR series and rehab devices. Control over the actuator layer gives the company more direct influence over weight, compliance, and reliability than a pure systems integrator.

Fourier is a vertically integrated robotics hardware company selling to B2B and institutional buyers, with the rehabilitation business providing a commercial base while the humanoid line moves toward broader deployment.

Its go-to-market is consultative across both segments. Rehab systems are sold to hospitals, outpatient clinics, and research centers through a solution-sales motion, including multi-device RehabHub deployments, joint lab arrangements, and clinical partnerships such as the Brooks Rehabilitation MOU in the US and the Kurage neurostimulation collaboration in France. Humanoids are sold to research institutions, universities, and early industrial or care-setting adopters, with GR-3 also supporting teleoperation-based industrial inspection and assembly use cases. Neither segment follows a self-serve or high-velocity sales model.

Revenue is captured primarily through capital equipment sales, with bundles and options layered on top, including dexterous hands, vision modules, custom head configurations, and software integration services. The open-source N1 and developer tooling appear aimed less at direct monetization than at ecosystem development: lowering adoption friction for labs and developers, expanding the install base around Fourier hardware, and keeping the platform relevant as embodied AI tooling becomes more standardized.

The business model relies on rehab revenue to fund humanoid R&D, while humanoid deployments produce the data and commercial proof points needed to expand into care and service robotics. Fourier's decade of building human-safe, body-interacting machines in clinical settings transfers into the safety and interaction design requirements for GR-3 and future care-oriented humanoids, capabilities that pure humanoid startups without a clinical heritage would need to develop separately.

The cost structure is capital-heavy relative to software businesses, with manufacturing complexity, component sourcing, and field support weighing on margins. Fourier's self-developed actuators provide some insulation from component cost dynamics that constrain Western competitors, but the business remains dependent on physical production scale to improve unit economics over time.

Fourier competes across two markets, rehabilitation robotics and general-purpose humanoids, with limited overlap outside care robotics. The competitive set, buying criteria, and evidence buyers use differ materially across the two.

The most direct pressure on Fourier's humanoid pricing comes from Unitree, which has set a price benchmark for low-cost developer humanoids at around $13,500 for entry-level units. Unitree's advantage is also its developer community: it sells humanoids as developer kits rather than enterprise procurements, which limits what Fourier can charge for research-oriented GR and N1 units.

Galbot, backed by Chinese industrial policy capital including CATL, Bosch, Toyota, and SAIC Motor, has secured several thousand unit orders and is scaling production faster than most Western competitors. Its industrial-and-consumer roadmap is similar to Fourier's sequencing but comes with more manufacturing velocity and heavier state support. Both companies benefit from Chinese supply-chain advantages in motors, actuators, and magnets, which give them a structural cost edge over US-based competitors, similar to the dynamic that let Chinese EVs undercut Western manufacturers.

Figure, Agility Robotics, and Apptronik are targeting industrial deployment and winning enterprise buyers with factory-specific operating data. Figure's BMW deployment, Agility's Toyota manufacturing agreement, and Apptronik's Jabil manufacturing partnership are the kind of industrial ROI references that Fourier's humanoid line has not yet matched at scale. For buyers focused on labor automation in controlled environments, those companies present a more direct value proposition than Fourier's care and research positioning.

1X is taking a different approach, focused on the home rather than the factory, and is the clearest Western example of the view that consumer environments generate the richest training data for general intelligence. Sunday Robotics and The Bot Company are pursuing non-humanoid form factors at sub-$10K price points, arguing that simplified wheeled platforms can reach commercial viability faster than full bipedal humanoids. If any of these approaches produces a robot that handles household tasks reliably at consumer price points before Fourier's GR-3 care-bot finds product-market fit, the care-humanoid niche Fourier is targeting becomes harder to defend.

In rehab, Fourier's most established rival is Hocoma, whose Lokomat, Andago, Erigo, and Armeo systems cover the continuum of neurological rehabilitation and are embedded in premium clinical workflows. Hocoma benefits from institutional inertia: therapists trained on its systems, procurement teams familiar with its evidence base, and referral patterns built around its product categories. Fourier's response is platform breadth through RehabHub and faster hardware iteration enabled by Chinese manufacturing, but displacing Hocoma in established neurorehab centers requires changing clinical habits as much as replacing equipment.

Ekso Bionics competes in FDA-cleared neurorehab exoskeletons with a clinician-first positioning that resonates in US procurement. Wandercraft is the most strategically relevant rehab competitor because it is pursuing a similar path, using rehab robotics as a path into humanoids, with its Atalante X exoskeleton deployed at 100+ rehab centers and a humanoid development program in parallel. If Wandercraft succeeds, it would validate Fourier's category thesis while creating a well-capitalized rival with gait-control IP and clinical credibility.

Fourier's expansion logic runs in two directions: deepening the clinical and care robotics market, where its rehab heritage provides a wedge, and broadening the humanoid platform into research, developer, and eventually general-purpose service markets.

GR-3's positioning as a care bot, with touch sensing, multimodal interaction, and a dual-path emotional response system, maps to a structural demand shift in global healthcare: aging populations in Japan, China, South Korea, and Europe are increasing pressure on care staffing capacity.

WHO projects that 1 in 6 people globally will be 60 or older by 2030, and the gap between care demand and available human labor is already visible in Japan and Germany. Fourier's rehab-to-care pipeline is more credible than that of a cold-start humanoid company entering the same space because it already has relationships with hospitals, eldercare institutions, and clinical researchers that understand workflow requirements. The GReAT Summit and partnerships with Singapore's National Healthcare Group and Nagoya University indicate this expansion path is moving from concept to implementation.

N1 and the FourierActionNet open-source dataset position Fourier as infrastructure for the embodied AI research community, competing with developer platforms at multiple price points, including Unitree's lower-cost alternatives. The open-source strategy is asymmetric: Fourier keeps proprietary leverage in actuator hardware and the core GR platform while opening enough of the software stack to attract universities, labs, and startups that generate use cases and training data.

As the embodied AI stack matures, with NVIDIA Isaac, simulation pipelines, and foundation model tooling becoming more accessible, the value of a well-documented, open-ecosystem humanoid platform increases. Developers who prototype on N1 are potential buyers of GR-2 or GR-3 for production deployments, and data generated across a broad research install base can feed back into Fourier's own model development.

Fourier's international footprint spans 40+ countries, with its Singapore office serving as the hub for Asia-Pacific expansion. Partnerships in Japan, Malaysia, and the broader Gulf region indicate geographic diversification, while the Brooks Rehabilitation MOU in the US and the Kurage partnership in France fit a clinical-credibility-first entry pattern before direct humanoid commercialization at scale in Western markets.

The BASF materials partnership points to deeper supply-chain integration that could support both cost reduction and product differentiation in markets where material quality and safety certification matter. That is particularly relevant for care-oriented humanoids entering European regulatory environments, where product-safety scrutiny for human-interactive robots is likely to be more demanding than in China.

Clinical proof gap: Fourier's care-bot ambitions with GR-3 require evidence of safety, workflow fit, and measurable outcomes in eldercare and hospital settings, a higher bar than a compelling demo, and one that rehab robotics incumbents like Hocoma and Ekso have spent years addressing through clinical trials, FDA clearances, and therapist training programs that Fourier has not yet replicated for its humanoid line.

Hardware commoditization: The same Chinese manufacturing ecosystem that gives Fourier a cost advantage over Western humanoid companies is also producing Unitree, Galbot, AgiBot, and a growing cohort of well-funded rivals that can drive down humanoid hardware prices faster than Fourier can build software and services moats, which could compress margins on the GR series before recurring revenue streams are large enough to offset that pressure.

Strategic dilution: Fourier is simultaneously running a mature rehab equipment business, a multi-generation humanoid R&D program, an open-source developer ecosystem, and a care-robotics commercial push, a breadth that risks leaving it outpaced by more focused competitors in each segment, whether Hocoma in clinical rehab, Figure in industrial humanoids, or Unitree in low-cost developer platforms.