1X and OpenAI Competitive Tension

This turns 1Xs cap table into a preview of where control in humanoids may ultimately sit, with the model supplier moving closer to becoming a full stack robot builder. OpenAI first backed 1X in its $23.5M Series A2 round, then later built out an in house robotics team with roles spanning actuators, control systems, field operations, and robot data collection. That means 1X is not just taking capital from a strategic investor, it is helping validate a future rival that wants access to the same ingredients, real world data, deployed workcells, and eventually distribution.1 2 3 4

• The tension is strongest around data. OpenAI Robotics explicitly describes a need for large live operating environments and deployed workcells to gather robot data, while 1X and other humanoid companies are racing to collect teleoperation logs and home interaction data that improve models over time.3 5 6
• This is common in AI infrastructure, but more sensitive in robotics because the investor can learn from the market while building its own product. Figure shows the adjacent version of this pattern, it has taken backing tied to AI leaders and is using industrial deployments at BMW to build its own embedded model and fleet learning loop.6 7
• 1X is also differentiated enough that the relationship still made sense. Its main push has been home robots, with NEO priced at $20K upfront or $499 per month, while most leading rivals have sequenced factories and warehouses first to generate uptime, revenue, and cleaner training data before going after the harder consumer market.5 6

The next phase is likely a split between robot makers that own the customer deployment and model builders that try to own the intelligence layer, with the biggest winners collapsing both into one stack. That makes 1Xs access to real homes and OpenAIs move into operating environments strategically important, because whoever captures the best physical world data earliest will shape the market.3 5 6