Gemini Robotics 1.5 and Gemini Robotics-ER 1.5
2025-09-25 · Publisher : Google · Domain : Artificial Intelligence
Google DeepMind launches Gemini Robotics 1.5, a vision-language-action model that thinks before acting, and Robotics-ER 1.5, a spatial reasoning model open to developers in preview through the Gemini API.
What changes
- Gemini Robotics 1.5: a VLA model that thinks before acting, explains its reasoning and learns across several types of robots.
- Gemini Robotics-ER 1.5: the first reasoning model optimized for embodied intelligence, which plans, calls tools (Google search, third-party functions) and drives a VLA.
- Best aggregate performance on 15 academic spatial reasoning benchmarks (ERQA, Point-Bench, etc.), according to Google.
- Adjustable thinking budget to trade off latency and accuracy.
- Update to the ASIMOV semantic safety benchmark.
### Specifications - Availability: ER 1.5 in preview in the Gemini API and Google AI Studio; Robotics 1.5 reserved for selected partners. - Exact API identifier not recorded in the sources consulted (the changelog mentions "Gemini Robotics ER 1.5" in preview). ### Pricing Pricing not disclosed in the announcement posts. ### Benchmarks Best aggregate performance on 15 academic benchmarks (score shown in a chart, not quantified in the text, scores published by the vendor). ### Key takeaways The first Gemini model for robotics offered to all developers, to build the planning "mind" of a robot.
Links
Tags : LLM, Google, DeepMind, Robotique, Gemini