Startup Eka secured a $1.45 billion valuation on 20 October 2026 after closing a $150 million Series A round led by venture funds a16z and F-Prime, according to TechCrunch. The company built a physical AI model and force-sensitive gripper that allows robots to handle dexterous manipulation tasks with speed.
Key takeaways
- Eka raised $150 million in a Series A round valuing the startup at $1.45 billion.
- The company developed a vision-force-action system instead of standard vision-language-action schemas.
- Eka's robotic arm can pick up about 1,500 objects in an hour.
- Founders Pulkit Agrawal and Tuomas Haarnoja used simulation to scale reinforcement learning.
How does Eka build its physical AI model?
Eka relies on a vision-force-action model instead of approaches tied primarily to visual inputs and text. Co-founder Pulkit Agrawal told TechCrunch that vision only gets a robot close, while actual grabbing happens through forces. The startup designed a flexible robotic hand with extra force sensors that reacts to contact rather than crushing delicate items like raspberries.
The underlying system uses a physics simulator to model multiple forces a robot encounters. This creates a reinforcement learning environment running at scale, generating thousands of hours of robot experience in minutes. University of Illinois Urbana-Champaign professor Kris Hauser noted that the robot adapts to many different objects without retraining.
What are the main robotics startup metrics?
TechCrunch reports that the Eka robot moves smoothly and rapidly, picking up about 1,500 objects weighing up to 1 kg in an hour, compared to roughly 600 for a human. It achieves a reliable pickup rate of over 95%. A16z general partner Raghu Raguram told TechCrunch that the system already exceeds superhuman performance for certain tasks.
Why it matters
The new architecture bridges the gap between general intelligence and physical execution that has troubled the robotics industry for years. While top-level planners can decide what tasks need doing, Eka provides the physical execution layer so robots can handle objects swiftly and safely without breaking them or failing on the final millimeter of contact.
Frequently asked questions
TechCrunch reports that Eka is valued at $1.45 billion post-money after closing a $150 million Series A funding round.
The startup was founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja.
The fundraising round was co-led by venture funds a16z and F-Prime.
Eka's robot can pick up about 1,500 objects weighing up to 1 kg in an hour, while maintaining a reliability rate of over 95%.
Eka uses a vision-force-action schema paired with a force-sensitive gripper, trained through physics simulation and reinforcement learning rather than human demonstration videos.
Based on reporting by TechCrunch. Read the original article
