Radar / AI robotics / Gemini Robotics 2.5
Gemini Robotics 2.5: DeepMind's most capable robot model, tested on a live Mercedes line
Google DeepMind introduced Gemini Robotics 2.5, its most capable vision-language-action model, reporting 30-100% higher success rates across 200+ tasks and a 96% success rate kitting car parts with Apptronik Apollo 2 humanoids on active Mercedes-Benz production lines.
Why it matters
A VLA model proven on a real factory floor, not a lab bench: humanoids sorting parts on live Mercedes lines, with faster skill learning and a new open robot-safety benchmark alongside.
What you could build with it
A warehouse automation team could prototype a kitting station around a vision-language-action model like this: a robot arm sorts mixed parts from bulk bins into ordered trays from a plain-language instruction, with a human supervisor approving each new part type before the robot runs it unsupervised.
Does it hold up?
Too early to judge for outsiders: the model is rolling out to DeepMind's robotics partners only, with no public API, price, or availability date — the Mercedes pilot is the only real-world evidence so far.
Built with Gemini Robotics 2.5
- Gemini Robotics 2.5 Put Humanoids on a Real Mercedes Line — 96% Success, and One Big Catcharticle · Independent read of the announcement: what the 96% figure does and does not prove, the 200-task evaluation, and the fact that nobody outside the partner group can call the model yet.
- Google DeepMind releases Gemini Robotics 2.5 to partners and an open ASIMOV-Jailbreak robot safety benchmarkarticle · Release coverage: the Mercedes proof of concept, 20-30% faster task completion, and the open ASIMOV-Jailbreak benchmark for testing whether robot AI can be tricked into unsafe actions.
- Apollo 2 Humanoids Hit 96% in Mercedes-Benz Sorting Proof of Conceptarticle · Pilot details: Apollo 2 robots on active lines in Kecskemét and Sindelfingen, the early-access waitlist, and Gemini 4 Argon's role in accelerating development through simulation.
Learn more
First spotted on article: source.
More AI robotics
RuView: see people through walls with WiFi, no camerasOpen-source Rust system that turns ordinary WiFi signals into real-time pose estimation, vital-sign monitoring, and…robotics · JEV 0.76Runway Praxis-1: open-weight robot model that learned to act from videoRunway released Praxis-1, a 27B-parameter world-action model, with open weights, inference and fine-tuning code under…robotics · JEV 0.66NVIDIA GR00T 2: a world-action robot model that learns from human videoNVIDIA's open frontier world-action model predicts future video and robot actions together, hitting 71% one-shot…robotics · JEV 0.66RoboParty RP1: a full-stack open-source humanoid robot for embodied AI researchRoboParty's 1.3-metre humanoid, launched 16 October and back in coverage on 2 November, comes in 24-DoF standard and…robotics · JEV 0.62T-Rex: a robot manipulation AI that reacts to fingertip touchPanasonic, UC Berkeley, NVIDIA and Stanford built T-Rex (Tactile-Reactive Dexterous Manipulation), a VLA model that…robotics · JEV 0.61Bear Robotics KINO-2: demonstration-to-skill robot-learning platformBear Robotics launched KINO-2 on Oct 21 at GTC Berlin: a robot-learning platform that turns human task demonstrations…robotics · JEV 0.59
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.