Editorial illustration for Google DeepMind unveils Gemini Robotics‑ER 1.6, beats prior model in tool count
DeepMind Gemini Robotics-ER 1.6: AI Tool Mastery
A robot that reaches for a ghost is a robot that fails. That cold fact drives every line of code in Google DeepMind’s latest release. Gemini Robotics-ER 1.6 just landed, and it does something its predecessor couldn't: count tools accurately.
Hammers. Scissors. Paintbrushes.
Pliers. Garden tools, each item tallied without error. The prior model missed scissors entirely, fumbled the count of paintbrushes, and hallucinated a wheelbarrow that simply wasn't there.
For robotics pipelines, that difference is the line between precision and catastrophe. A hallucinated object isn't just a bug; it's a robot lunging at empty space. Now the new model also reads multiple views and knows when a task is done, two capabilities that turn perception into reliable action.
Gemini Robotics-ER 1.6 doesn’t just count tools better. It stops seeing ghosts. That’s the real leap.
For a robot, hallucination isn’t a quirky mistake; it’s a collision course with nothingness. ER 1.5 reaches for a wheelbarrow that isn’t there. ER 1.6 sees the scene as it is, no empty grasps, no wasted motion.
This is the difference between a machine that fumbles through uncertainty and one that acts with grounded confidence. The upgrade matters most where it’s hardest to measure: in the quiet moment when a task ends. Knowing when to stop is the final frontier of embodied reasoning.
ER 1.6 takes that step. It reads the room, counts what counts, and leaves the phantom tools to the prior generation. Physical AI is only as reliable as its perception.
Now it’s finally seeing clearly.
Related Reading
Browse all Google news →
Research & Benchmarks Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks Databricks finds multi-step agents beat single-turn RAG by 21% to 38% on STaRK
Research & Benchmarks Nvidia's DLSS 4.5 beta adds 6x Multi Frame Generation for RTX 50 GPUs
Related Topic NVIDIA and Google Cloud let developers scale AI from prototype to production
Market Trends
Related Topic Google tests visual 'magazine-style' UI for Gemini 3 Pro users
LLMs & Generative AI
From Archives UK tests Mythos AI, noting its ability to chain multistep attacks
From Archives AI Forum Launches Professional Certificate and USD 120M Fund for AI Fluency
From Archives Google adds “Skills” to Chrome, enabling one‑click reuse of Gemini prompts
LLMs & Generative AI
From Archives Google leaders, Hassabis, refute uneven AI adoption, cite 40K SWEs agentic coding
Market Trends
Common Questions Answered
How does Gemini Robotics-ER 1.6 improve tool identification compared to its previous version?
Gemini Robotics-ER 1.6 demonstrates superior tool identification capabilities by correctly counting hammers, scissors, paintbrushes, pliers, and garden tools in a scene. Unlike its predecessor, the new model avoids hallucinating objects that are not present and provides more accurate visual recognition of multiple tool types.
What external capabilities does the Gemini Robotics-ER 1.6 system possess?
The Gemini Robotics-ER 1.6 can invoke external utilities like Google Search and vision-based APIs, indicating an advanced level of autonomous tool interaction. This feature positions the system as a potential 'cognitive brain' for robots, enabling more sophisticated task planning and reasoning capabilities.
What are the key improvements in embodied reasoning for the Gemini Robotics-ER 1.6?
The new model promises enhanced embodied reasoning and instrument reading, allowing robots to better understand and manipulate real-world objects. It demonstrates improved spatial reasoning and task detection, with the ability to accurately identify and count various tools without mistaking absent objects for present ones.
Further Reading
- Gemini Robotics ER 1.6: Enhanced Embodied Reasoning — Google DeepMind Blog
- Gemini Robotics-ER 1.6 - Google Blog — Google Blog
- Gemini Robotics-ER 1.6 - Google DeepMind — Google DeepMind
- Google Unveils Enhanced Gemini Robotics Model (GOOG) — GuruFocus
- Google DeepMind Releases New AI Model to Bring Robots Closer to Real Autonomy — TipRanks
