Eka Robotics Raises $150M Series A at $1.45B for a Vision-Force-Action Model Trained in Simulation
Eka Robotics closed a $150M Series A at a $1.45B post-money valuation led by a16z and F-Prime, TechCrunch reported. a16z calls Eka's approach a Vision-Force-Action model trained mainly by reinforcement learning in physics simulation. Pick-rate figures are laboratory results: a16z cites up to 1,500 picks per hour in lab induction and bin-picking tasks.
Physical-AI investors are still paying unicorn prices for force-aware manipulation, even in a week when AI-infrastructure financing looked strained. Eka Robotics, founded by MIT professor Pulkit Agrawal and former Google DeepMind researcher Tuomas Haarnoja, closed a $150 million Series A at a $1.45 billion post-money valuation led by a16z and F-Prime, TechCrunch reported on Tuesday, October 20.
Instead of a vision-language-action stack, Eka pairs a force-sensitive gripper with a model trained mainly by reinforcement learning in physics simulation. a16z — the lead investor — calls it a Vision-Force-Action model. a16z partners Yoko Li and Raghu Raghuram wrote the firm's investment note.
Lab performance claims
TechCrunch says the robot picks about 1,500 objects of up to 1 kg an hour, against about 600 for a human, and that Eka says it succeeds more than 95% of the time. a16z describes "up to 1,500 picks per hour in laboratory induction and bin-picking tasks," about 2.5× a human. Those are lab figures, not production-line results.
The gripper costs under $1,000 and is rated for 5 million duty cycles, against a 20 million target for industrial arms, TechCrunch reports. Pilots with industrial customers in pick-and-place and wire handling are under way but undisclosed.
Limits
- a16z is the lead investor; its post is not an independent check.
- Pick rates and reliability are company or laboratory claims.
- Field uptime and task generality beyond pick-and-place remain unproven in the coverage we read.