Google DeepMind releases Gemini Robotics 2.5
What happened
On 10 November 2026, Google DeepMind released Gemini Robotics 2.5, a vision-language-action (VLA) model it calls its most capable VLA model to date. According to the release information, the model improves success rates by 30% to 100% over the previous baseline across 200 tasks and completes tasks 20% to 30% faster. On the same day, Ars Technica reported that DeepMind released version 2.5 of the Gemini Robotics platform, saying it shows clear gains on tasks requiring high dexterity and was evaluated across more than 200 activities, including serving food with tongs and using a pipette; the report also mentioned the open-source safety benchmark ASIMOV-jailbreak released around the same time. The original release information described the evaluation as "200 tasks", while later reporting said "more than 200 activities", a slight difference in wording. On 11 November, The Decoder reported that on Mercedes-Benz production lines in Kecskemét, Hungary, and Sindelfingen, Germany, Apptronik's Apollo 2 humanoid robot used the model to sort metal parts and place them in trays, with a pilot success rate of 96%. There is currently no independent third-party evaluation.
Written by AI from the reports · updated 5 d ago
Timeline
Follow the reports to see the event from each side.
- The DecoderSelectedGoogle DeepMind's Gemini Robotics 2.5 scores 96 percent sorting metal parts in a Mercedes-Benz assembly pilot
Google Deepmind introduced Gemini Robotics 2.5, a new control model for robots, and tested it with Apptronik on active Mercedes-Benz production lines in Hungary and Germany. The humanoid robot Apollo 2 sorted metal parts there with a reported success rate of 96 percent.
- Ars Technica · AIDeepMind reveals Gemini Robotics 2.5 and open safety benchmark "ASIMOV-jailbreak"
The new model is up to twice as successful at some tasks.
- Google DeepMindSelectedGemini Robotics 2.5 accelerates real-world general dexterity
Our most capable vision-language-action model enhances robot dexterity, reliability, and safety
Heat of this event
Current heat 3·Peak in the comparable range 22(Nov 11 05:00)·Change over 24 hours in the comparable range -80%
The trend compares only the same accounts observed without a break, so its range can be narrower than the current heat count. Move the pointer or tap the chart to see each hour; with a keyboard, use the left and right arrow keys.