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

The first is that the data constraint is being attacked with simulation rather than teleoperation. Lucida, published on 1 September, converts ordinary video of cluttered rooms into editable assets for robot simulators. Tripo AI raised about $446 million for generative 3D two months after a $150 million round, with simulation demand behind it.

The second is that foundation models have started to work in the control loop. NavMCP reported a 78.3 percent success rate on a Unitree Go2 by scaffolding a vision-language model onto a navigation foundation model, with double-digit gains over prior benchmarks. MIT's Phillip Isola has proposed cloud-hosted language models controlling connected robots directly.

The third is that the compute is being contracted. Figure AI signed a $3.5 billion agreement with Nscale targeting 100,000 Nvidia GPUs, deploying in the second half of 2027.

A new entrant raising at $1.2 billion before shipping is raising into that.

The uncomfortable comparison

Figure's $3.5 billion compute deal is nearly three times XDOF's entire proposed valuation, and Figure has been at this for years with humanoids in customer trials.

That is the structural problem for a young robotics company: the incumbents are not other startups, they are startups with balance sheets large enough to buy a hundred thousand accelerators on a two-year forward commitment. Catching up on the data-and-compute axis is not something a $1.2 billion Series B funds.

Which implies XDOF's investors are betting on a different axis — a specific form factor, a specific industrial application, or an architectural claim that reduces how much data is needed. XDOF has not publicly described which.

What is not known

Almost everything. XDOF has not disclosed what it builds, who its customers are, whether it has a product in trials, who is leading the round, or whether terms are agreed. The talks are reported, not closed.

The honest summary is that a company three months out of stealth is being valued at $1.2 billion on the strength of the category and the founders. That is a defensible thing for a venture investor to do in a market forming this fast. It is also how the last several rounds of robotics enthusiasm began, and those ended with capable machines and no economics.