Google DeepMind

A vision-language-action (VLA) model is one neural network that reads camera images and a plain-language instruction and outputs a robot's next motor commands, built by training a vision-language model on robot demonstrations. Google DeepMind, Physical Intelligence, NVIDIA, Figure AI, Skild AI and AgiBot lead, and the commercial race now turns on the cost of action data rather than model design.

Runway announced Praxis-1 on 30 September 2026, an open-weight world action model that converts its large-scale video pretraining into control for real robots. Early tests run with Noble Machines, Standard Bots and Ultra on bimanual, six-axis arm and mobile-base hardware, and public weights are promised in the coming months.

Shipments of professional service robots rose 24% to almost 250,000 units in 2025, led by 117,500 transport and logistics robots, the International Federation of Robotics said on 30 September 2026. The IFR counted about 7,000 full-size humanoids sold worldwide and 174 humanoid makers, more than half of them in China, and said most humanoid uses still rely on teleoperation.

General Intuition, which trains AI models on action-labelled gameplay clips from the Medal platform, said on 29 September 2026 that it had raised $220 million at a $6.2 billion valuation from investors including Khosla Ventures and General Catalyst. The figure is 2.7 times the $2.3 billion it disclosed with its Series A 96 days earlier, and the company has opened a waitlist for partners in robotics, simulation and games.

Hitachi has reached a basic agreement with Munich's Agile Robots to fuse their AI models into a robot "brain" for factory work and pair it with Hitachi's edge AI chips, then sell the results globally through HMAX Industry. No financial terms or launch dates were disclosed, and the deal came a day before Hitachi's separate physical AI partnership with FANUC.

Extend Robotics said on Sep 28 that it closed a GBP 2.6 million Pre-A round led by Skyworks Venture Capital Fund, with Neo Venture (Europe) and Zip Capital joining. The Reading-based teleoperation company charges customers for completed work rather than equipment, says it has 35 subscription customers, and will use the money to expand across UK and European industry and build a remote operator network.

Generalist AI said on August 19, 2026 that GEN-1.5 can learn a new manipulation task from one demonstration of 3 to 12 seconds placed in its context, averaging 59% success over 10 tasks with no fine-tuning and 83% after 10 gradient steps on about 5 minutes of data. The company says the ability emerged from more than eight months of pretraining, but calls the tasks simple and the skills brittle, and has named no customers.

Google DeepMind on July 30 released Gemini Robotics 2, a family of three models covering whole-body humanoid control, high-level embodied reasoning and on-device inference. The reasoning model ER 2 is available to developers through the Gemini API and Google AI Studio, while the action models remain limited to early-access partners and trusted testers.

NVIDIA announced at GTC Taipei on June 1, 2026 an Isaac GR00T reference humanoid for academic research, built on Unitree's H2 Plus body with two 22-degree-of-freedom Sharpa Wave tactile hands and a Jetson AGX Thor T5000 computer. Unitree will sell it from late 2026; Ai2, ETH Zurich, Stanford and UC San Diego are the first named research users. No price was given.

FANUC said on May 13 that all of its robots will be fully supported on Intrinsic's enterprise software platform and that it has joined Google DeepMind's Gemini Robotics trusted tester programme, after shipping more than 1,000 robots for physical AI uses since December. Two days later it expanded its NVIDIA work, linking ROBOGUIDE with Isaac Sim, training a dual-arm robot with Isaac GR00T N and moving a collision-avoidance system to Jetson Thor.

Sony AI's autonomous table tennis system Ace won three of five matches against elite players and lost both matches against two Japanese T.League professionals in April 2025 trials, according to a paper published in Nature on April 22, 2026. The robot combines event-based vision sensors from Sony Semiconductor Solutions, an eight-axis arm and reinforcement learning policies trained in simulation, and reaches a perception latency of 10.2 milliseconds.

An 82-author team from Toyota Research Institute, Cornell University and MIT reported in Science Robotics on April 15, 2026 that multitask large behavior models, built on diffusion policy, outperform single-task baselines in blind, randomized trials and need far less task-specific data. The preprint version describes about 1,700 hours of robot demonstrations across more than 500 tasks and 1,800 controlled real-world trials.

Google DeepMind released Gemini Robotics-ER 1.6 on April 14, 2026, an embodied reasoning model built on Gemini 3.0 Flash and offered as a preview through the Gemini API and Google AI Studio. DeepMind said the model reads industrial instruments correctly 86% of the time, or 93% with an agentic zoom and code execution step, against 23% for ER 1.5. Boston Dynamics is using the capability for Spot facility inspections.

Generalist AI released GEN-1 on April 2, 2026, an embodied foundation model pretrained on more than 500,000 hours of human physical interaction data captured with wearable devices and no robot data. The company reports 99% average success on tasks where its previous model reached 64%, about one hour of robot data per task and task times roughly three times faster than the prior state of the art; the results are the company's own and have not been independently replicated.

Fauna Robotics said on 24 March 2026 that it had joined Amazon and would operate as an Amazon company, keeping its Sprout humanoid on sale and its customers supported; no terms were disclosed. The deal gives the operator of more than 1 million warehouse robots a small, safety-first humanoid used by Disney, Boston Dynamics, NYU and UC San Diego, and a team drawn from Google DeepMind and CTRL-labs.

Munich-based Agile Robots said on 24 March 2026 it had formed a strategic research partnership with Google DeepMind to integrate Gemini Robotics foundation models with its hardware, starting with manufacturing use cases. The German company, which says it has installed more than 20,000 robotic solutions, becomes the latest hardware maker in Google's orbit after Boston Dynamics and Apptronik. No financial terms were disclosed.

NVIDIA used its GTC keynote on March 16, 2026 to preview Isaac GR00T N2, a robot foundation model built on its DreamZero world action model research that it says succeeds at new tasks in new environments more than twice as often as leading VLA models. It also put the 3-billion-parameter GR00T N1.7 into commercial early access and announced Cosmos 3 and Isaac Lab 3.0, with N2 due by the end of 2026.

Sunday, the Silicon Valley developer of the Memo household robot, said on 12 March 2026 that it had raised an oversubscribed $165 million Series B at a $1.15 billion valuation, led by Coatue, with Coatue's Thomas Laffont joining the board. The company trains Memo on data from people wearing its Skill Capture Glove instead of teleoperating robots, plans an invite-only home beta from the autumn and expects to ship at scale in 2027-2028.

Hyundai Motor Group agreed on February 27, 2026 to invest about 9 trillion won from 2026 in a robotics, AI and hydrogen hub at Saemangeum in North Jeolla. The plan includes a 400 billion won robot cluster with a finished-robot factory sized for about 30,000 units a year, a contract-manufacturing foundry and a parts complex, due in 2029, next to a 5.8 trillion won AI data centre scaled in phases to 50,000 GPUs.

Austin humanoid maker Apptronik said on 11 February 2026 that it had closed a $520 million Series A-X extension priced at three times its Series A valuation, lifting total Series A funding above $935 million. John Deere, AT&T Ventures and Qatar Investment Authority joined returning backers B Capital, Google, Mercedes-Benz and PEAK6, and the company said the money will pay for Apollo production, training and data facilities and a new robot due in 2026.

Figure AI on Jan. 27, 2026 released Helix 02, a robot foundation model that runs a humanoid from camera pixels and fingertip touch down to joint commands at 1 kHz, replacing 109,504 lines of hand-written C++ with a 10-million-parameter learned controller. The company showed a Figure 03 unloading and reloading a dishwasher across a full kitchen in a four-minute, 61-action run it says had no resets and no teleoperation.

Microsoft Research on January 21, 2026 introduced Rho-alpha, its first robotics model, derived from the Phi vision-language family and trained partly on synthetic data from NVIDIA Isaac Sim on Azure. The model adds tactile sensing and learning from operator corrections to the usual vision-language-action recipe, and is offered through a research early access program ahead of a later Microsoft Foundry release.

Hyundai Mobis will supply actuators for the production version of Boston Dynamics' Atlas humanoid under a framework announced at CES on January 7, 2026, its first robotics customer. The Hyundai parts maker says actuators make up more than 60% of a humanoid's material cost, and plans to extend the tie-up to grippers, sensors, controllers and battery packs as Hyundai Motor Group targets 30,000 robots a year by 2028.

Boston Dynamics unveiled the production version of its electric Atlas humanoid at CES 2026 on January 5 and said manufacturing begins immediately in Boston, with all 2026 units already committed to Hyundai Motor Group and Google DeepMind. Hyundai set a goal of a production system able to build 30,000 robots a year by 2028, when Atlas is due to start parts sequencing at its Georgia Metaplant, moving into assembly by 2030. Hyundai Mobis will supply the actuators and Google DeepMind's Gemini Robotics models are to supply the cognition.

Robot foundation model developer Physical Intelligence has raised $600 million at a $5.6 billion valuation in a Series B led by Alphabet's growth fund CapitalG, according to its investors. The round followed the release of π*0.6, a model the company says more than doubles throughput on hard tasks after reinforcement learning on real robots.

Generalist AI introduced GEN-0 on November 4, 2025, a family of embodied foundation models that the company says improves predictably with more physical interaction data. It reported a capacity threshold at 7 billion parameters, models scaled past 10 billion, and a pretraining corpus of more than 270,000 hours growing by over 10,000 hours a week. The findings were published as a company blog post, not a peer-reviewed paper.

Google DeepMind on 25 September 2025 introduced Gemini Robotics 1.5, a vision-language-action model that reasons in natural language before moving and transfers skills across robot bodies, alongside Gemini Robotics-ER 1.5, an embodied reasoning model opened to developers as a preview in the Gemini API. In its technical report, the pairing of the two cut total subtask failures on long-horizon tasks to 22% from 44.5% when a general Gemini 2.5 Flash model did the planning.

Skild AI on July 29, 2025 named and described its robotics foundation model, the Skild Brain, a two-layer system pre-trained on simulation and internet video that the company says runs across quadrupeds, humanoids, table-top arms and mobile manipulators. The launch carried no customers or prices, but set out the body-agnostic strategy that later underpinned a $14 billion-plus valuation and more than $100 million in annual recurring revenue.

Genesis AI, a physical AI lab based in Paris and Palo Alto, came out of stealth on July 1, 2025 with $105 million co-led by Eclipse and Khosla Ventures, with Bpifrance, HSG, Eric Schmidt and Xavier Niel also investing. The company, which later called the round seed funding, plans a universal robotics foundation model trained on a data engine that mixes in-house physics simulation, generative modelling and real robot data, and did not disclose a valuation.

Google DeepMind on June 24, 2025 released Gemini Robotics On-Device, a vision-language-action model built to run locally on a robot with no network link. It is the first VLA the lab is letting outside developers fine-tune, through a new Gemini Robotics SDK offered to trusted testers, and DeepMind says it adapted the model from its ALOHA training rig to a bi-arm Franka FR3 and Apptronik's Apollo humanoid.

Physical Intelligence on 22 April 2025 published π0.5, a vision-language-action model that drove mobile manipulators through 10- to 15-minute cleaning jobs in three homes absent from its training set. The San Francisco start-up says co-training on data from other robots, web tasks and spoken coaching, rather than more home-robot data alone, is what lets the model generalise, and its own ablations show which ingredients matter most.

NVIDIA on 18 March 2025 released Isaac GR00T N1, a dual-system vision-language-action model for humanoid robots, as downloadable weights on Hugging Face and GitHub, with 1X, Agility Robotics, Boston Dynamics, Mentee Robotics and NEURA Robotics named as early-access partners. It also announced Newton, an open-source physics engine under development with Google DeepMind and Disney Research, a synthetic manipulation-data blueprint and a 15-terabyte open physical AI dataset.

Google DeepMind on 12 March 2025 introduced Gemini Robotics, a vision-language-action model built on Gemini 2.0 that issues motor commands, and Gemini Robotics-ER, a spatial reasoning model for robot makers. The lab said the action model more than doubles performance on a generalization benchmark against other leading models, named Apptronik as its humanoid partner and opened Gemini Robotics-ER to four trusted testers including Boston Dynamics and Agility Robotics.

Figure AI on 20 February 2025 introduced Helix, a vision-language-action model that pairs a 7-billion-parameter vision-language model running at 7 to 9 Hz with an 80-million-parameter control policy running at 200 Hz across 35 degrees of freedom. Figure says the model was trained on about 500 hours of teleoperation data and runs entirely on two low-power embedded GPUs inside the robot.

Austin humanoid maker Apptronik said on 13 February 2025 it had raised a US$350 million Series A co-led by B Capital and Capital Factory, with Google participating. The round is more than 12 times the US$28 million the company had raised before and will fund Apollo production, new robot iterations and wider deployments in 2025.