# Eka Robotics Hits $1.45B Valuation With Physical AI Breakthrough A novel vision-force-action model enables robots to achieve superhuman dexterity in object manipulation. - **Source:** [TechCrunch](https://techcrunch.com) - **Published:** 2026-10-20 - **Topic:** [Software & Engineering](https://brifing.si/c/engineering) - **Original article:** - **On Brifing:** https://brifing.si/s/98bf2c63-9318-427b-9255-24098ded8a7c ## Unicorn Status Achieved Eka Robotics secured a $150 million Series A funding round, pushing its post-money valuation to $1.45 billion. Led by a16z and F-Prime, the investment highlights significant confidence in the startup's unique manipulation architecture. Founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja, the firm aims to redefine robotic dexterity. ## Beyond Vision-Language Models Unlike standard vision-language-action models, Eka utilizes a vision-force-action schema. This approach prioritizes the relationship between language, force inputs, and physical actions, addressing the critical 'last millimeter' problem in robotics. By focusing on tactile feedback rather than just visual planning, the system achieves precise control necessary for handling delicate objects without damage. ## Simulation-Driven Reinforcement Learning The company bypasses the need for massive datasets from human demonstrations by using physics simulators. This environment allows the robot to generate thousands of hours of experience in mere minutes. By training through reinforcement learning at scale, the system learns to surpass human-provided examples, enabling it to adapt to novel objects without retraining. ## Superhuman Manipulation Metrics Eka’s robot demonstrates impressive speed and reliability, picking up 1,500 objects per hour with a 95% success rate. This significantly outperforms human capabilities, which typically hover around 600 objects per hour. The system's ability to handle items weighing up to 1kg showcases a robust integration of hardware and software control algorithms. ## Tactile Sensitivity in Action The hardware includes a specialized, flexible gripper equipped with advanced force sensors. This design allows the robot to interact with fragile items, such as raspberries, with extreme gentleness. By reacting to contact in real-time, the system avoids the crushing force common in traditional industrial grippers, demonstrating a sophisticated level of physical awareness. ## Bottom-Up Intelligence Strategy While competitors focus on high-level planning, Eka adopts a bottom-up approach to robotics. The team believes that solving physical interaction first is essential for broader intelligence. By handling the physicality of tasks, their model complements general-purpose AI, which can then provide high-level decision-making without needing to manage the complex mechanics of movement. ## Industrial Deployment Challenges Eka is currently conducting secretive pilots with industrial customers focusing on pick-and-place and wire manipulation tasks. The primary challenge remains scaling these lab results to factory floors. The hardware must prove it can operate reliably for millions of cycles, with the current goal being to reach 20 million duty cycles for industrial standards. ## Future Scaling Potential The company is prioritizing pragmatic applications to prove its value in specific verticals before expanding. By mastering pick, arrange, place, and wire handling, Eka aims to establish a repeatable engineering workflow. Success in these initial deployments will determine the company's ability to bridge the gap between specialized tasks and general-purpose robotic automation. ## Sources covering this story - https://techcrunch.com/2026/10/20/eka-is-worth-1-45-billion-thanks-to-a-physical-ai-model-that-has-experts-raving/ — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._