Google DeepMind
Build with Gemini Try Gemini

Gemini Robotics On-Device

Gemini Robotics 1.5

Powering an era of physical agents to transform how robots actively understand their environments

Join the trusted tester waitlist Try Gemini Robotics-ER 1.5 on AIS

Gemini Robotics models allow robots of any shape and size to perceive, reason, use tools and interact with humans. They can solve a wide range of complex real-world tasks – even those they haven’t been trained to complete.

Read the blog

Gemini Robotics 1.5 is designed to reason through multi-step complex tasks, and to make decisions to form a plan of action. It will then work to carry out each step autonomously.

Capabilities

Gemini models are capable of responding to text, images, audio, and video. Gemini Robotics adds the ability to reason about physical spaces – allowing robots to take action in the real world.

  • Generality

    Understands the physical world, and adapts and generalizes its behaviour to fit new situations. Breaks down goals into manageable steps to make longer-term plans and overcome unexpected problems.

  • Agentic

    Assess complex challenges, natively call tools – like Google Search – to look up information, and create detailed step-by-step plans to overcome them.

  • Thinking

    Enables robots to think before acting, improving the quality of their actions, and making their decisions more transparent in natural language.

  • Interactivity

    Understands and responds to everyday commands. Can explain its approach while taking action. Users can redirect it at any point, without using technical language. It also adjusts to any changes in its environment.

  • Dexterity

    Enables robots to tackle complex tasks requiring fine motor skills and precise manipulation – like folding origami, packing a lunch box, or preparing a salad.

  • Multiple embodiments

    Adapts to a diverse array of robot forms, from bi-arm static robotic platforms like ALOHA and Bi-arm Franka, to humanoid robots like Apptronik’s Apollo. A single model can be used across all these robots, in turn accelerating its learning across multiple embodiments.

Hands-on

See how Gemini Robotics handles a range of different tasks.

Agentic capabilities

Uses digital tools autonomously to solve complex tasks.

Watch

Thinking while acting

Solves longer, multi-step tasks – without needing new instructions after each step.

Watch

Learning across embodiments

Transfers learned motions across robots of different sizes and shapes, helping robots to become more useful.

Watch

Embodied reasoning

Understand its environment and how to complete a task.

Watch

Generality in action

Generalizes across novel situations and solves a vast range of tasks.

Watch

Dynamic interactions

Responds to natural conversation and adapts rapidly to changing environments.

Watch

Partnering with Apptronik

Helping to build the next generation of humanoid robots.

Watch

Dexterous skills

Perform tasks that require fine motor skills and coordination.

Watch

Responsibly advancing AI and robotics

To ensure Gemini Robotics benefits humanity, we’ve taken a comprehensive approach to safety, from practical safeguards to collaborations with experts, policymakers, and our Responsibility and Safety Council.

Learn more

Model and tools

We take a dual-model approach, pairing a vision-language-action (VLA) with an embodied reasoning (ER) model. Each model plays a specialized role, working together as a powerful and versatile system.

Gemini Robotics SDK

Helping developers easily adapt our Gemini Robotics On-Device model to new tasks and environments.

Sign up to access the SDK

Collaborations

We’re partnering with Apptronik to build the next generation of humanoid robots. We’re also working with over 60 trusted testers to guide the future of Gemini Robotics-ER.

Experience
Gemini Robotics

If you're interested in testing our models, please share a few details to join the waitlist.

Join the trusted tester waitlist
Gemini Robotics On-Device

Gemini Robotics On-Device

Our vision-language-action (VLA) model that’s optimized to run locally on robotic devices

The general-purpose dexterity and task adaptation capabilities of Gemini Robotics, optimized to run efficiently on-device.

Read the blog

Capabilities

Incredibly versatile, leveraging the intelligence and state-of-the-art manipulation capabilities of our Gemini Robotics model.

  • Dexterity

    Designed for rapid experimentation with dexterous manipulation capabilities.

  • Adaptability

    Fast adaptation to new tasks and environments, after learning from fewer than 100 examples.

  • Runs locally

    Optimized to run locally on robots, with low latency inference.

Gemini Robotics On-Device is the first VLA model we're making available for robotics developers to fine-tune. This means they’ll be able to adapt the model to achieve better performance for their applications.

Try Gemini Robotics SDK

Model information

View model card

Model deployment status Private preview
Supported data types for input Image, Text, Action
Supported data types for output Action
Supported # tokens for input 1088
Availability Trusted tester