Apptronik leans on Google DeepMind's Gemini to give Apollo a brain
The Austin robotics maker is building its next humanoids on DeepMind's Gemini Robotics models — and Google is now an investor too.
AUSTIN — Apptronik is building the next generation of its Apollo humanoid on Google DeepMind's Gemini Robotics models, a partnership that pairs the Austin company's hardware with one of the most advanced foundation-model efforts aimed at the physical world. Google is also among the lead backers of Apptronik's recent $520 million funding extension.
The collaboration matters because the hardest part of humanoid robotics is no longer the body — it is the brain: the perception, planning and dexterity required to operate in messy, unstructured spaces. Gemini Robotics is DeepMind's attempt to bring large-model reasoning to manipulation tasks, and Apollo gives it a commercially aimed platform already being tested in Mercedes-Benz factories and GXO warehouses.
Why it matters
The humanoid race is splitting into hardware companies, model companies, and the few trying to do both. Apptronik's choice to lean on DeepMind for intelligence — while owning the robot and the customer relationships — is a bet that the model layer will commoditize faster than the integration and deployment layer. If that thesis holds, the Austin company's value lives in getting real robots working at real customers, not in training the underlying brain itself.
Apptronik is also working with Google DeepMind to build the next generation of humanoid robots, powered by Gemini Robotics models.— Apptronik, company materials
Reported by Next in Austin. Based on reporting from SiliconANGLE.
