Google to Replace Gemini’s Gems with Skills in November

Google is planning a big change to its Gemini AI app. Starting November 17, 2026, the company will retire a feature called “Gems” and replace it with something new called “Skills.” This means the custom AI assistants users have built with Gems will be transformed automatically into Skills, so no one has to start from scratch.

Gems were introduced in 2024 as a way for users to create personalised AI helpers for specific tasks. You could make a Gem that acted as a running coach, a career adviser, or even a coding partner. These Gems had their own instructions, could hold files for reference, and even shared with others. It was a handy tool for people wanting a custom AI experience.

But Google is moving towards using Skills inside Gemini Spark, a different part of the app that focuses on managing tasks and workflows. Skills are also custom instructions, but they work a bit differently. Instead of launching separate chatbots like Gems, Skills can be called up during any Spark chat by typing a slash “/” command. They can also work in the background to help with tasks and be combined for more complex jobs.

One important difference is that Skills are currently only available to paying subscribers of Google AI Pro or Ultra plans. Gems, on the other hand, have been free to use. This shift could mean that some users who enjoyed free Gems might have to pay to keep using their AI helpers once the migration happens.

Users will still be able to use their Gems until the November 17 deadline. After that, Google will automatically convert them into Skills. However, some details about how the migration will work exactly, and if all features like attached files and sharing links will carry over, are still unclear. Google hasn’t shared full information on this yet.

This move comes as other companies like Meta push forward with new all-in-one AI assistants, such as their Muse agent, which can handle many tasks across apps in one place. Google seems to be adjusting its strategy to focus more on integrated task management rather than separate AI personalities.