Company profile
Google DeepMind
Google DeepMind develops artificial intelligence research and applied AI systems. Its robotics work includes vision-language-action models, robot learning,...
Facts
- Segment
- FOUNDATION_MODELS
- Country
- GB
- Headquarters
- London
- Founded
- 2010
- Categories
- Software
- Website
- https://deepmind.google
About Google DeepMind
Google DeepMind develops AI models and research systems for reasoning, perception and physical-world interaction. Its current robotics work centres on Gemini Robotics vision-language-action models, which convert visual information and instructions into motor commands across robot arms and humanoid platforms.
The wider research lineage includes RT-2 for vision-language-action control and AutoRT for directing robots to collect training data using foundation models. DeepMind also provides MuJoCo simulation and embodied-reasoning models for spatial understanding, planning and robot application development. Availability varies across research releases, trusted-tester access and private-preview programmes.
Market context
DEVELOPMENT
Summary
Type: Robot AI
Core: Gemini Robotics
Function: Reasoning
Audience: Developers
Research, Articles and recent verified events
- Google DeepMind selects 15 startups for European Robotics Accelerator — 2026-06-09
- Innate runs 25 MARS robots at Y Combinator RoboHacks — 2026-04-11
- Agile Robots and Google DeepMind form strategic research partnership for industrial robotics — 2026-03-24
- Google DeepMind launches robotics accelerator for European startups — 2026-03-04
- Google DeepMind hires former Boston Dynamics CTO Aaron Saunders for robotics hardware push — 2025-11-19
- Google DeepMind launches Gemini Robotics 1.5 for agentic physical tasks — 2025-09-25
- Google DeepMind introduces Gemini Robotics On-Device for local robot control — 2025-06-24
- Google DeepMind introduces Gemini Robotics and Gemini Robotics-ER — 2025-03-12