Google DeepMind releases Gemini Robotics 2.5 vision-language-action model

Advances general-purpose robot dexterity with a frontier VLA model and a new safety benchmark.

Gemini Robotics 2.5 is DeepMind's most capable vision-language-action model, aimed at improving robot dexterity, reliability and safety. It is released to partners alongside an open ASIMOV jailbreak safety benchmark, and scored 96% sorting Mercedes-Benz metal parts in a pilot.

Name

Gemini Robotics 2.5

Domain

industry

Read it at Google DeepMind →

Coverage: 18 sources, in order

1. 09 Nov 19:51 Google DeepMind Read the article →
2. 10 Nov 15:00 Google DeepMind Read the article →
3. 10 Nov 15:30 Reddit Read the article →
4. 10 Nov 16:54 MLQ.ai Read the article →
5. 10 Nov 17:11 The Decoder Read the article →
6. 10 Nov 17:11 The Decoder Read the article →
7. 10 Nov 17:43 Unite.AI Read the article →
8. 10 Nov 18:16 Firstpost Read the article →
9. 10 Nov 19:15 NewsBytes Read the article →
10. 10 Nov 20:42 Interesting Engineering Read the article →
11. 10 Nov 20:42 Interesting Engineering Read the article →
12. 11 Nov 06:00 Tech in Asia Read the article →
13. 11 Nov 07:15 Tech Times Read the article →
14. 11 Nov 08:45 India Today Read the article →
15. 11 Nov 11:54 Dataconomy Read the article →
16. 11 Nov 13:54 AI Insider Read the article →
17. 11 Nov 15:31 www.tokenpost.com Read the article →
18. 11 Nov 16:30 SiliconANGLE Read the article →

← Front page