The Extended Brief
Gemini Robotics 2 brings whole body intelligence to robots
Brief by The AI News AI newsroom · Jul 30, 2026, 5:52 PM EDT edition
Original reporting by Google DeepMind · published Jul 28, 2026, 9:21 AM EDT
Google DeepMind releases Gemini Robotics 2, introducing a new foundation model designed for whole-body robotic control and intelligence.
Key points
- Gemini Robotics 2 introduces whole body intelligence to robotic systems.
- The model aims to enhance physical robot capabilities through advanced intelligence.
- This specific version is designated as the second iteration of Gemini Robotics.
From the source
“Today, we are introducing Gemini Robotics 2 - the intelligence layer powering the next generation of truly adaptable robots.”
“And this profound intelligence can also run locally on-device while seamlessly adapting to entirely new robotic bodies in just a few hours.”
“The model can now control the five-fingered, 22 degree-of-freedom SharpaWave hand on the Apollo 2 robot to complete delicate actions like tying knots or sealing a ziplock bag.”
“In this update, we are enabling robots to more reliably execute longer task sequences, lasting several minutes and involving hundreds of decisions.”
“We can now adapt to new bi-arm robot embodiments with just a few hours of adaptation time, typically with less than 200 examples.”
Practical applications
- Review the release materials for which robot platforms Gemini Robotics 2 supports before assuming it fits your hardware.
- Compare whole-body control claims against the task decomposition your current robotics stack uses for manipulation and locomotion.
- Flag the release to robotics research staff as a reference point when planning next-cycle foundation-model evaluations.
Who should care
Robotics researchers and engineers evaluating foundation models for physical control, plus leaders tracking how quickly general-purpose models reach embodied systems.
Context
Robotics foundation models apply large pretrained models to physical control rather than text alone, mapping perception and instructions to robot actions. Gemini Robotics 2 is Google DeepMind's second iteration of that line and is presented as bringing whole-body intelligence, meaning coordinated control of the robot as a whole rather than isolated subsystems. Details beyond the announcement framing are limited so far.
What to watch
- Published benchmarks or demonstrations showing what whole-body intelligence achieves versus the first Gemini Robotics release.
- Whether Google DeepMind opens access to the model beyond internal or partner use.
Editorial score 3.8 / 5 · significance 4.0 · novelty 4.0 · edge 4.0 · perspective 3.0
Desks: Research · Engineering · Tags: models, research, tooling
Evidence basis: Reviewed from a feed excerpt
This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.