Flexion, a robotics intelligence company, announced a $50 million Series A round from DST Global Partners, NVentures, redalpine, Prosus Ventures, and Moonfire, following a recent $7.35 million seed raise. The new capital accelerates development of a reinforcement learning and sim-to-real platform designed to power adaptable humanoid robots across tasks and environments.
Flexion builds a full autonomy stack consisting of a command layer for language-based reasoning, a motion layer trained on synthetic and real-world edge-case data, and a control layer enabling low-latency whole-body actions. This architecture supports deployment with minimal human involvement, replacing scripted logic and tele-operation with adaptive learning systems.
The robotics sector continues to face a gap between mechanical capability and true intelligent autonomy. Flexion targets this missing intelligence layer, aiming to transform hardware into scalable, useful systems capable of operating in real-world conditions. As global labor shortages intensify and industries seek automation beyond controlled settings, humanoid robots represent an economic necessity rather than a distant vision.
Record compute advances now enable robotics to reach the kind of utility inflection point seen in large language models. Flexion’s new funding will support expansion of its Zurich R&D center, increased compute and robot fleets, a U.S. presence, and broader commercialization efforts. Existing collaborations with major OEM partners will also scale globally.
Flexion’s team brings experience from ETH Zurich, NVIDIA, Meta, Google, Tesla, and Amazon, with backgrounds in reinforcement learning, control systems, perception, and simulation. The company aims to deliver the intelligence stack that allows humanoid robots to work alongside humans.
"Our mission is simple but ambitious: to power the intelligence stack for humanoid robots, so they can work alongside humans, not depend on them. We’re not building the body. We’re building the brain. This is a hard problem. But the right kind of hard."