← Leaderboard

Gemini Robotics 1.5

VLA
Google DeepMind · Sep 25, 2025

Multi-embodiment VLA that interleaves actions with natural-language reasoning (“thinking before acting”) and uses a Motion Transfer mechanism so skills trained on one robot work on the others. Evaluation is self-reported on DeepMind’s own task suites; the model is available only to select partners.

Type
VLA
Parameters
—
Backbone
–
Weights
Closed
License
–
Commercial use
Unclear
Released
Sep 25, 2025
Robots
ALOHA 2, bi-arm Franka, Apptronik Apollo
SO-101
–
LeRobot
–
Inputs
–
VRAM
–

Benchmark results

0 of 13 benchmarks

No benchmark results recorded yet.

Not reported on 13 benchmarks
Wrong or missing number? Report an error

Built on Gemini Robotics 1.5