Google is changing how users create and access customized experiences inside Gemini. The company is moving away from its Gems feature and transitioning those custom assistants into a new format called skills.
The change is expected to take effect on November 17, 2026. Google says existing Gems will be migrated automatically, meaning users who have already created custom assistants will not need to recreate them manually. Until the transition date, existing Gems are expected to remain available.
From Custom Assistants to Skills
Google introduced Gems in 2024 as a way for users to create personalized versions of Gemini for specific tasks.
Instead of repeatedly entering the same detailed instructions, users could create a Gem with predefined guidance and return to it whenever they needed help with a particular workflow.
Google’s original examples included tools for learning, brainstorming, career guidance, writing and coding. Users could also create their own specialized assistants for activities such as planning trips, improving fitness routines or handling recurring personal tasks.
Google later expanded Gems with sharing capabilities, allowing users to make customized assistants available to friends, family members or colleagues.
Now, Google is folding that concept into the broader idea of skills.
Users Don’t Have to Rebuild Their Work
One important part of the transition is that existing custom configurations are not simply being removed.
Google says Gems will automatically be converted into skills. This means users who have invested time developing customized instructions should not have to start over.
The approach also reflects a wider shift in Google’s AI strategy toward reusable capabilities that can be applied across different workflows rather than keeping specialized functions separated behind individual product names.
Google has already been using the concept of skills in its developer-focused agent ecosystem. Its Gemini API documentation describes agents that can be configured with their own instructions, skills and data.
A Different Way to Use Customized AI
The transition could change how users interact with their personalized AI tools.
According to the reported Gemini interface, users will be able to call a skill within a task by entering a forward slash and selecting the relevant capability.
That approach puts more emphasis on choosing a capability inside an ongoing conversation rather than opening a separate assistant designed around a particular role.
For experienced AI users, this can make specialized functions feel more integrated into a broader workflow. For casual users, however, the extra interaction step could make the feature less immediately obvious.
Why Google Is Consolidating Features
The move also highlights a recurring challenge in the fast-moving AI market: product features can evolve quickly, while their names and interfaces change along with them.
Gems were originally positioned as a distinct Gemini experience for personalized AI assistance. The new skills model places similar functionality within a broader system of reusable capabilities.
That could make Google’s AI ecosystem more unified, particularly as Gemini evolves from a conventional chatbot toward an environment that can perform more complex tasks.
Google has been expanding its agent capabilities across its ecosystem. In 2026, the company introduced managed agents for the Gemini API that can reason, use tools and execute code within isolated environments.
What the Change Means for Gemini Users
For people already using Gems, the immediate takeaway is relatively straightforward: existing custom assistants are being transitioned rather than simply erased.
The bigger change is conceptual.
Instead of thinking about personalized AI as a collection of individually named assistants, Google appears to be moving toward a model where specialized capabilities can be activated when they are needed within a larger AI workflow.
As AI platforms increasingly become capable of handling multiple tasks, reusable skills could become an important part of how users organize their interactions with intelligent systems.
The November transition will show how smoothly Google’s new approach works in practice and whether users find skills easier to discover and use than the previous Gems experience.
