Skip to content

Eka Robotics

Request correction

Corporate brand

Unclaimed

American robotics company building a Vision-Force-Action foundation model to give robots general, high-performance, safe physical-world intelligence.

Last updated September 28, 2026

AI crawler reads

No AI crawler has read this profile yet.

Overview

Eka Robotics is an American robotics company focused on building artificial intelligence for the physical world. Rather than treating robotics as a hardware-first discipline, the company positions itself around a foundation-model approach it calls Vision-Force-Action (VFA). According to its official website, the VFA model is intended to resolve a long-standing trade-off in robotics between generality and performance, uniting generality, performance and safety so that robots can move beyond human limits and become accessible to a broad set of users. The company frames its technology around three claimed capabilities. "General" mastery means a single model that generalizes across objects, tasks and environments. "Performant" mastery means fast, reliable robots that adapt to the unexpected. "Interactive" mastery means safe, collaborative, environment-agnostic operation alongside humans. Eka Robotics describes itself as building intelligence for the physical world "in its native language: force," and it markets its work under the banner of "The Era of Superhuman Robotics." Eka Robotics presents itself as a research-driven venture. Its website states that the team is made up of pioneers of self-supervised learning, large-scale robot learning, world models, reinforcement learning and sim-to-real robotics, and that it is "now deploying intelligence into the physical world." The company lists affiliations with MIT, Berkeley, Harvard, CMU, BU, UPenn, DeepMind, Microsoft and Boston Dynamics as sources of its talent, and it actively recruits through a "Join Us" call to action. Public contact is provided through a general email address and LinkedIn and X (Twitter) channels. As of the available evidence, Eka Robotics is an early-stage, privately held company. It is described in structured public data as an American robotics company. No founding year, founder names, headquarters city, funding history, product pricing or customer list are established by the supplied sources, and the company's public footprint at the time of writing consists mainly of its own website and a Wikidata entity. Its copyright notice reads "© 2026 EKA Robotics, Inc.," indicating the legal entity name EKA Robotics, Inc.

History

Eka Robotics is an American robotics company whose public record, as reflected in the supplied evidence, is brief and recent. The company is identified in structured public data as an American robotics company, and its own website presents it as a venture building artificial intelligence for the physical world. The available sources do not establish a founding date, founding team, headquarters location, or funding history, so its early corporate history cannot be reconstructed from the evidence at hand. What the company does communicate is a research and technology thesis. Eka Robotics argues that robotics has until now forced a choice between generality and performance, and that its Vision-Force-Action (VFA) model resolves that trade-off by uniting generality, performance and safety. The company describes this as building intelligence for the physical world "in its native language: force," and it frames the moment as "The Era of Superhuman Robotics." Its stated goal is to push robots beyond human limits and into everyone's hands. The company organizes its claims around three capability pillars. "General" denotes mastery of the unknown through a single model that generalizes across objects, tasks and environments. "Performant" denotes mastery of physics through fast, reliable robots that adapt to the unexpected. "Interactive" denotes mastery alongside humans through safe, collaborative, environment-agnostic operation. Together these pillars describe a foundation-model approach rather than a single hardware product. Eka Robotics also emphasizes the provenance of its team. Its website states that the company was "built by the team that makes the impossible possible," describing its people as pioneers of self-supervised learning, large-scale robot learning, world models, reinforcement learning and sim-to-real robotics, and stating that it is now deploying intelligence into the physical world. The company lists MIT, Berkeley, Harvard, CMU, BU, UPenn, DeepMind, Microsoft and Boston Dynamics among the institutions and employers its team comes from, and says that list is growing. It recruits publicly through a "Join Us" call to action and provides a general contact email plus LinkedIn and X channels. The company's copyright notice reads "© 2026 EKA Robotics, Inc.," which indicates the legal entity name EKA Robotics, Inc. and suggests the site content is current as of that year. Beyond this, the supplied evidence does not document product launches, commercial deployments, partnerships, funding rounds, leadership appointments or other dated milestones. As a result, the history presented here is limited to what the company itself asserts about its mission, technology and team, and the entry should be treated as an early-stage profile rather than a complete corporate history.

Products and positioning

Research-driven robotics AI company building a general, high-performance and safe foundation model for physical-world intelligence.

  • Vision-Force-Action (VFA) model

    A foundation model for physical-world intelligence that the company says unites generality, performance and safety, generalizing across objects, tasks and environments while adapting to the unexpected and operating safely alongside humans.

Recent events

  • Unconfirmed

    Reported

    Eka Robotics closes $150 million Series A[3]

    Eka Robotics closed a $150 million Series A round that valued the startup at $1.45 billion post-money. The round was led by a16z and F-Prime.

Read more

Sources4 sources

Cite this profile: Cite the canonical profile. /brand-wiki/eka-robotics · Editorial policy · How profiles are compiled