


OpenAI is working on a new "Prompts" section where users will be able to store their prompts with template variables and tools being predefined. Business plans only?

By
–



OpenAI is working on a new "Prompts" section where users will be able to store their prompts with template variables and tools being predefined. Business plans only?
By
–
Gemini CLI can now run complex and interactive commands directly within your Gemini CLI session! Top, Vim, and others.
— π¨ AI News | TestingCatalog (@testingcatalog) 16 octobre 2025
CLIMINI π€ https://t.co/8fLaLIntm6 pic.twitter.com/aWa5UQMIhc
Gemini CLI can now run complex and interactive commands directly within your Gemini CLI session! Top, Vim, and others. CLIMINI

By
–
Google AI Studio is rolling out a new model selector for its Build section! As well as a microphone selector and a new prompt page.

By
–
Image, video and audio modes are now available via model selection. UI revamp is complete!
By
–
BREAKING π¨: Google is working on Tools support for custom Gems! It is yet unclear which tools will be supported, in fact, but the current list is wild.
— π¨ AI News | TestingCatalog (@testingcatalog) 16 octobre 2025
Upcoming Canvas Creative and Visual Layout tools are currently in the Labs as well. pic.twitter.com/KdLgybaowJ
BREAKING : Google is working on Tools support for custom Gems! It is yet unclear which tools will be supported, in fact, but the current list is wild. Upcoming Canvas Creative and Visual Layout tools are currently in the Labs as well.

By
–
Microsoft is starting to roll out Connectors on Copilot! Currently available connectors include apps from Microsoft and Google Suites. h/t @h_a_r_r_i_s_pl
By
–
NotebookLM now supports LaTeX formulas! π€ https://t.co/MdxDEOSDVs pic.twitter.com/JJn646s4n6
— π¨ AI News | TestingCatalog (@testingcatalog) 16 octobre 2025
NotebookLM now supports LaTeX formulas!

By
–
ICYMI: Besides Veo 3.1, Google Flow editor got a new video in-painting feature. Also, Frames and Ingridients are now using Veo 3.1

By
–

Sora Pro users now can access Storyboard feature and generate up to 25 seconds videos on the web. h/t @koltregaskes
By
–
BREAKING π¨: Google is preparing Gemini Agent prototype to perform your tasks on the web!
— π¨ AI News | TestingCatalog (@testingcatalog) 15 octobre 2025
It is likely to be powered by the recently released Computer User model and will be able to operate a browser and perform deep research.
With great power comes responsibility π€ pic.twitter.com/LcXnphOVEL
BREAKING : Google is preparing Gemini Agent prototype to perform your tasks on the web! It is likely to be powered by the recently released Computer User model and will be able to operate a browser and perform deep research. With great power comes responsibility