Pomegra
Eka Robotics closed a $150M Series A led by a16z and F-Prime at a $1.45B valuation, betting robot hands should rely on force rather than vision alone.
Key Takeaways
- Eka Robotics raised $150M in a Series A at a $1.45B post-money valuation, announced October 20, 2026.
- a16z and F-Prime Capital led the round, about 17 months after an $11M seed.
- Eka claims 1,500 picks per hour at over 95% success, against roughly 600 for humans.
Lead
Eka Robotics, a Cambridge, Massachusetts startup, announced a $150M Series A on October 20, 2026, valuing the company at $1.45B post-money. Andreessen Horowitz (a16z) and F-Prime Capital co-led the round. Eka builds a force-sensitive gripper and a "vision-force-action" model, an architecture that treats touch as the primary signal for manipulation. The company has been public for roughly a year and a half, and its seed round was only $11M.
Who Is Backing Eka Robotics?
Eka's Series A drew a mix of new and returning backers, with a16z and F-Prime in the lead seats. Seed leads Obvious Ventures and E14 Fund returned. Also named are MFV Partners, Bison Ventures, Glasswing Ventures, Samsung Next and a group of unnamed Fortune 500 strategic investors.
The seed round, co-led by Obvious and E14 in May 2025, raised $11M. A jump from $11M to $150M in about 17 months is steep even by current robotics standards, and the post-money figure implies investors are paying for the founding team and early benchmarks rather than commercial revenue. Eka has not disclosed revenue or customer contracts.
What Does Eka Robotics Actually Build?
Eka builds a custom gripper and a vision-force-action (VFA) model that together control robot manipulation. Most competitors train vision-language-action models, which infer movement mainly from camera input. Eka's argument, in the words of co-founder Pulkit Agrawal, is that vision gives a rough location while "the actual grabbing happens through forces."
The company trains its models with reinforcement learning in physics simulation, which it says can generate thousands of hours of robot experience in minutes. Those skills are then transferred to physical hardware. The gripper costs under $1,000 to build and is rated for 5 million duty cycles, with a stated goal of 20 million.
How Well Does It Perform?
Eka says its system picks about 1,500 unfamiliar objects weighing under 1 kg per hour, with a success rate above 95%. It puts the human benchmark at roughly 600 per hour. Those figures come from the company, and no independent test has been published.
Two verticals are in focus: pick-and-place and wire handling. Pilots with industrial customers are underway, though the company has not named them. The product is in an early field-deployment phase.
Who Founded Eka?
Eka was founded by Pulkit Agrawal, an MIT associate professor and co-founder of SafelyYou, and Tuomas Haarnoja, a former Google DeepMind researcher who co-created the Soft Actor-Critic algorithm. The two met as graduate students at UC Berkeley. The team draws on alumni of MIT, Berkeley, DeepMind, Boston Dynamics, Microsoft, Carnegie Mellon, the University of Pennsylvania, Northeastern and Oxford.
How Does the Valuation Compare With Rivals?
At $1.45B, Eka sits well below the leaders in robot foundation models. Physical Intelligence is valued at $11B, Generalist at $3B and Skild AI at more than $14B. Eka's price is roughly 13% of Physical Intelligence's.
That gap cuts two ways. It leaves room for a markup if pilots convert into contracts. It also reflects that Eka is narrower, with a specific hardware bet on touch and a smaller set of target tasks than the generalist models its peers are pursuing.
What Comes Next for Eka?
Eka plans to use the money to scale VFA model capabilities, widen the range of tasks its robots handle and speed up customer deployments. The test will be whether simulation-trained force control holds up on messy factory and warehouse floors, where objects, lighting and wear vary daily.
If pilots become paid deployments, the sub-$1,000 gripper and the throughput claim give Eka a clear cost argument against manual labor. If they stall, the lead investors will have paid a unicorn price for a thesis that remained unproven outside the lab.
Outlook
Eka Robotics now has $150M, a $1.45B valuation and top-tier backers behind a contrarian view that force, not sight, should drive robot manipulation. The performance numbers are promising but self-reported, and customers remain unnamed. The next 12 months of pilot conversions will show whether the premium over its $11M seed holds.
