Identity
- Status
- preview
- Developer ID
gemini-robotics-2- Released
- Jul 30, 2026
- Version
- 2
- License
- Unknown
Google DeepMindGemini Robotics 2
Google DeepMind's private-preview vision-language-action model for tabletop, bimanual, and whole-body robot control.
- Model type
- Vision-language-action model
- Inputs
- Text, Image, Robot state
- Inference
- Unknown
- Released
- Jul 30, 2026
Providers & Pricing
0 current hosted offerings
No Current Hosted API Offering Verified
No exact current provider endpoint or published price has been verified for this model version.
At a Glance
Model Facts
Model Design
- Model type
- Vision-language-action model
- Inputs
- Text, Image, Robot state
- Outputs
- Robot action
- Total parameters
- Unknown
- Architecture
- Vision-language-action model
Deployment
- Inference
- Unknown
- Control rate
- Unknown
- API available
- Unknown
- Open weights
- No
- Self-hostable
- Unknown
Publisher Media
Gemini Robotics 2 Demonstrations
Embodied Intelligence
Robotics Control Profile
Control Contract
Model typeVision-language-action model
Action representationMotor-control actions
Control architectureVision-language-action model
InferenceUnknown
Native control rateUnknown
Training Data and Embodiments
Not disclosed in the cited model page.
Verified Platform Links
Compatible Robots
| Robot | Relationship | Form | Availability | Evidence note | Source |
|---|---|---|---|---|---|
 Apollo 2Apptronik | Official demonstration | Humanoid | Select customers | Google DeepMind reports whole-body and dexterous manipulation results on Apollo 2 configurations. | Official fact |
This page represents one developer model product. Dated API snapshots, serving endpoints, pricing, regions, quantizations, and service tiers attach as versioned aliases or provider details and never create duplicate public model records.
Version Lineage
PredecessorUnknown
SuccessorsUnknown
Aliasesgemini-robotics-2
Explicit Unknowns
- Native control rate
- Total parameters
- License
- API availability
Primary Evidence
Sources and Observation Date
Comparable Peers
Related Model Comparisons
| A | Pair | B | Context |
|---|---|---|---|
Gemini Robotics 2Google DeepMind | vs |  MolmoAct 2Ai2 | cross-developer peersrobotics |
Gemini Robotics 2Google DeepMind | vs |  Helix 02Figure | cross-developer peersrobotics |
Gemini Robotics 2Google DeepMind | vs |  SmolVLA 450MHugging Face LeRobot | cross-developer peersrobotics |
Gemini Robotics 2Google DeepMind | vs |  GR00T N1.7 3BNVIDIA | cross-developer peersrobotics |
Gemini Robotics 2Google DeepMind | vs |  OpenVLA 7BOpenVLA Research Team | cross-developer peersrobotics |
Gemini Robotics 2Google DeepMind | vs |  pi 0.7Physical Intelligence | cross-developer peersrobotics |
Questions
Gemini Robotics 2 FAQs
What is Gemini Robotics 2?+
Google DeepMind's private-preview vision-language-action model for tabletop, bimanual, and whole-body robot control. It is developed by Google DeepMind and its current sourced lifecycle status is preview.
What inputs and outputs does Gemini Robotics 2 support?+
Gemini Robotics 2's current record lists Text, Image, and Robot state as input and Robot action as output.
What kind of robotics model is Gemini Robotics 2?+
Gemini Robotics 2 is classified as a vision-language-action model. Its action representation is Motor-control actions, its control architecture is Vision-language-action model, and inference is recorded as unknown.
Which robots and embodiments work with Gemini Robotics 2?+
The developer reports Apptronik Apollo 2 with Inspire hands, Apptronik Apollo 2 with Sharpa hands, and Franka Duo with Robotiq gripper as supported or represented embodiments. Model Markets also has sourced model-to-platform links for Apollo 2; each link states whether it is native, demonstrated, used for training, or used for evaluation.
How large is Gemini Robotics 2?+
The current record lists an unknown total parameter count and an unknown active parameter count. Its architecture is Vision-language-action model.
What capabilities and tasks does Gemini Robotics 2 support?+
Gemini Robotics 2's recorded capabilities are cross-embodiment, dexterous-manipulation, whole-body-control, and multi-robot-collaboration. Its supported tasks are robot-control, dexterous-manipulation, and whole-body-control.
What training data is reported for Gemini Robotics 2?+
Not disclosed in the cited model page.
Is Gemini Robotics 2 available through an API?+
Gemini Robotics 2's current primary-source record does not establish API availability. No provider endpoint is treated as available without direct provider evidence.
How much does Gemini Robotics 2 cost through an API?+
No directly sourced token price is currently attached to Gemini Robotics 2. Missing prices remain unknown rather than being inferred from another model or provider.
Can Gemini Robotics 2 be self-hosted?+
Gemini Robotics 2 is not marked as an open-weight model. Self-hosting is not established by the current source, with no license recorded.