Google DeepMind Adds Agentic Capabilities to AI Models for Robots

Google Deepmind

Google DeepMind introduced two artificial intelligence models to help developers build robots that can understand their environment and perform complex tasks.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    The new models build upon the Gemini Robotics models the company introduced in March by adding advanced thinking that enables agentic experiences, according to a Thursday (Sept. 25) blog post.

    The new Gemini Robotics 1.5 is a vision-language-action (VLA) model that turns visual information and instructions into motor commands, while the new Gemini Robotics-ER 1.5 is a vision-language model (VLM) that creates multistep plans to complete a mission, the post said.

    Gemini Robotics-ER 1.5 was made available to developers Thursday, while Gemini Robotics 1.5 is offered to only select partners, per the post.

    Carolina Parada, senior engineering manager at Google AI, said in the post that these models mark “a foundational step toward building robots that can navigate the complexities of the physical world with intelligence and dexterity.”

    “Gemini Robotics 1.5 marks an important milestone toward solving AGI in the physical world,” Parada said. “By introducing agentic capabilities, we’re moving beyond models that react to commands and creating systems that can truly reason, plan, actively use tools and generalize.”

    Advertisement: Scroll to Continue

    PYMNTS reported in March that robots are now in fashion in Silicon Valley, as large language models are giving robots the capability to understand natural language commands and do complex tasks.

    In addition to Google DeepMind’s Gemini Robotics, Meta’s PARTNR, Nvidia’s Isaac Groot N1, Tesla’s Optimus and a slew of AI robotics startups like Figure AI and Cobot are developing humanoid robots that can do general tasks.

    FieldAI said Aug. 20 that it raised $405 million in two consecutive rounds to accelerate the global adoption of its general-purpose robots. The company said it is developing a single software brain that can power a variety of robots, and its robots are currently operating on a day-to-day basis in construction, manufacturing, urban delivery and inspection.

    In July, Skild AI introduced an AI model it said can run on almost any robot. The company said its Skild Brain lets different kinds of robots think, function and respond more like humans.

    For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.