XDOF, just three months out of stealth, is in talks for a Series B at a $1.2B valuation

The world of artificial intelligence and robotics is moving at an incredible pace, and a recent development highlights just how quickly foundational infrastructure for these fields is gaining traction. XDOF, a startup specializing in robot training data, is reportedly in advanced discussions to secure a Series B funding round that could value the company at a staggering $1.2 billion. This news is particularly striking as it comes merely three months after XDOF officially emerged from stealth mode.

XDOF's Rapid Ascent in the AI Landscape

XDOF, pronounced "ecks-doff," made its public debut in June 2026, announcing an initial funding round of $70 million. This initial investment was backed by prominent venture capital firms including Thrive Capital, Spark Capital, a16z (Andreessen Horowitz), Lux, and WndrCo. The company was founded in October 2024 by a team of former UC Berkeley researchers: Philippe Wu, who serves as CEO, Fred Shentu as CTO, and Nemo Jin as COO.

The core mission of XDOF is to address one of the most significant challenges in modern robotics: the scarcity of high-quality data needed to train sophisticated robot models. Unlike large language models that can leverage vast amounts of publicly available text, robots require data that captures complex physical interactions, such as grasping objects, navigating environments, and performing intricate manipulation tasks. Such data is hard to come by, and existing video footage is often too low-fidelity to be directly useful for training robots.

What XDOF Brings to the Table

XDOF positions itself as a crucial infrastructure partner for robotics builders, providing the tools, data, and services necessary to accelerate the development of physical AI. The company has developed a comprehensive ecosystem designed to generate, clean, and annotate the specific data robots need to learn and operate effectively in the real world.

Their approach involves a three-tier data collection system:

• Target Robot Teleoperation Data: This involves directly controlling the exact robot being deployed to capture precise operational data.
• General Teleoperation Device Collection Data: Utilizing specialized devices like GELLO, which was co-developed by XDOF's founders during their time at Berkeley, to generate broader teleoperated data.
• Human First-Person Task Data: XDOF plans to develop its own wearable sensors to capture human-centric data, offering a unique perspective on how humans perform tasks.

Beyond data collection, XDOF also provides essential services like data cleaning, toolchains, and annotation systems, recognizing that raw data alone is insufficient. Many leading AI labs prefer to outsource these labor-intensive and specialized tasks, making XDOF's offering highly valuable.

As a testament to its capabilities, XDOF collaborated with the University of California, Berkeley's AI Research Lab to release the ABC robot training dataset (ABC-130K). This dataset is touted as one of the largest high-quality robot manipulation datasets available, featuring 130,000 trajectories, 300 hours of simulation data, and 100 hours of evaluation data. This data has already been used to train robots for complex tasks like folding T-shirts and loading AirPods into their cases.

Since its public launch, XDOF has quickly amassed approximately 20 clients, including several unnamed "frontier AI labs" that are at the forefront of robotics research and deployment. The company currently employs about 60 people, showcasing rapid growth in its operational capacity.

The Significance of a $1.2 Billion Valuation

The reported talks for a Series B round at a $1.2 billion valuation, just three months after its initial funding announcement, underscore several key trends in the AI and robotics sectors:

• Intense Demand for Robot Data Infrastructure: The speed at which XDOF is seeking and potentially securing this valuation highlights the urgent and unmet need for specialized data infrastructure in robotics. As major AI labs, including those that previously scaled back robotics efforts (like OpenAI, which relaunched its robotics program in Q2 2026), increasingly focus on physical AI, the bottleneck isn't just compute power or model architecture, but the data that enables real-world physical interaction.
• Investor Confidence in Foundational AI: Investors are recognizing that the next wave of AI innovation, particularly in robotics, requires robust foundational layers. Companies like XDOF, which provide the "picks and shovels" for this new gold rush, are seen as critical enablers for the broader industry. The significant jump in valuation from its initial $70 million funding to a potential $1.2 billion in mere months reflects a strong belief in XDOF's market position and execution.
• The Rise of Physical AI: The broader market for physical AI startups is experiencing a boom. According to PitchBook data, physical AI startups raised a record $16.3 billion across 492 deals in the first quarter of 2026. This surge is driven by falling hardware costs, labor shortages, and a renewed focus on reshoring manufacturing, all of which necessitate more capable and autonomous robots.
• Strategic Outsourcing by AI Labs: Building and maintaining the extensive infrastructure required for high-fidelity robot data collection, cleaning, and annotation is a massive undertaking. It demands significant warehouse space, hundreds of robots, and specialized personnel for maintenance, calibration, and operation. Many leading AI labs find it more efficient and cost-effective to outsource these complex tasks to specialists like XDOF.

This rapid valuation increase positions XDOF as a key player in the emerging ecosystem of physical AI. It suggests that the company's strategy of focusing on the often "dirty and unglamorous" but essential work of robot data collection is paying off significantly.

Future Outlook

With a potential Series B funding round at such a high valuation, XDOF would be well-capitalized to further expand its operations, accelerate research and development into new data collection methods (such as its planned wearable sensors), and potentially scale its client base. The capital would likely be used to enhance its data pipelines, develop more sophisticated collection tools and annotation systems, and grow its teams of teleoperators and egocentric data operators. This investment would solidify XDOF's role as a critical enabler for companies and labs aiming to deploy more capable and general-purpose robots in various industries.

As the robotics industry continues its accelerated growth, fueled by advancements in AI and increasing practical applications, companies that provide fundamental infrastructure like XDOF are poised for substantial impact. The reported Series B talks at a $1.2 billion valuation are not just a win for XDOF, but a strong indicator of the immense market potential and investor appetite for the foundational technologies driving the next generation of intelligent machines.

Frequently Asked Questions

What is XDOF and what problem does it solve?

XDOF is a robot data infrastructure startup founded in October 2024. It addresses the critical shortage of high-quality training data needed to develop and train general-purpose robots. Robots require complex physical interaction data, which is difficult and expensive to collect, clean, and annotate. XDOF provides the tools, systems, and services to overcome this bottleneck.

When did XDOF exit stealth mode and what was its initial funding?

XDOF emerged from stealth mode in June 2026, announcing an initial funding round of $70 million. This round included investments from Thrive Capital, Spark Capital, a16z, Lux, and WndrCo.

What makes XDOF's technology unique for robot training data?

XDOF employs a three-tier data collection system, including direct robot teleoperation, general teleoperation devices, and future plans for human first-person task data using wearable sensors. They also offer comprehensive data cleaning, tooling, and annotation services, and have co-released a large, high-quality robot manipulation dataset called ABC-130K with UC Berkeley.

What is the significance of XDOF's reported $1.2 billion valuation?

The reported $1.2 billion Series B valuation, coming just three months after exiting stealth, signifies strong investor confidence in XDOF's solution to a critical industry problem. It highlights the urgent demand for robust robot data infrastructure and the burgeoning growth of the physical AI and robotics sectors.