Google plans to retire Gems on Gemini by October 20

Google is reportedly preparing to retire Gems, the custom AI assistants available within Gemini, by October 20, 2026. Users are expected to be asked to save their existing Gem instructions or recreate them using a newer feature called Skills, signaling a shift in how Google wants people to customize and extend Gemini. The reported change was shared by TestingCatalog, which said users would be prompted to migrate their Gems to Skills.

The move would represent another step in Google’s broader effort to evolve Gemini from a chatbot with customizable personas into a more capable AI platform built around reusable workflows and task-specific capabilities. Gems were introduced as a way for users to create customized versions of Gemini with specific instructions, while Skills appear to be positioned as a more structured mechanism for defining how Gemini performs particular tasks.

What Happened

Google is reportedly planning to retire Gems in Gemini on October 20, 2026.

According to information shared by TestingCatalog, Gemini users will be asked to preserve their existing Gem configurations or recreate them as Skills before the retirement date.

Google has not yet published a detailed official announcement confirming the October 20 date in the public Gemini release notes reviewed for this report. Google’s public Gemini release-notes page continues to document updates to the service, but does not currently provide detailed information about the reported Gems retirement.

That means the reported date should be treated as a planned transition rather than a finalized public product announcement until Google provides additional documentation.

What Are Gemini Gems?

Gems were introduced by Google as customizable versions of Gemini.

They allow users to provide specific instructions that shape how Gemini responds or behaves. Instead of repeatedly giving the same instructions in a normal conversation, users can create a Gem designed for a particular purpose.

For example, a user could configure a Gem to act as:

• A writing assistant
• A coding tutor
• A study partner
• A research assistant
• A fitness-planning assistant
• A business-analysis helper
• A language-learning partner

The concept allows users to create a persistent set of instructions around a recurring task.

This was an important step in Google’s attempt to make Gemini more personalized and useful for repeated workflows.

Why Google Is Moving Toward Skills

The reported transition from Gems to Skills suggests Google may be moving away from custom AI personalities toward more task-oriented capabilities.

A Gem is primarily defined by instructions about how Gemini should behave.

A Skill can potentially represent a repeatable capability or workflow that Gemini can apply when needed.

That distinction becomes increasingly important as AI assistants evolve from answering questions to performing multi-step tasks.

Instead of simply telling Gemini to “act like a financial analyst,” for example, a Skill could potentially define a structured process for analyzing a financial report, extracting key figures, comparing periods and producing a standardized output.

The exact capabilities of Skills and how they will replace Gems have not yet been fully detailed publicly.

Gems vs Skills

Feature	Gems	Skills
Primary concept	Customized Gemini assistant	Reusable capability or workflow
Main purpose	Personalized responses	Task-oriented behavior
User instructions	Central to configuration	Expected to remain important
Reusability	Yes	Expected to emphasize repeatable tasks
Reported status	Planned retirement	Expected replacement
Transition date	October 20, 2026, reportedly	Migration target
Official detailed migration guide	Not yet publicly confirmed	Awaiting further details

The distinction is important because Google appears to be changing not only the name of the feature but potentially the underlying philosophy of customization.

What Users May Need to Do

If Google’s reported transition goes ahead, users with existing Gems will need to review their configurations before the retirement date.

The reported migration process would involve saving the instructions associated with existing Gems and recreating them as Skills.

Users should therefore consider preserving:

• Custom instructions
• Prompt templates
• Reference material
• Workflow descriptions
• Preferred response formats
• Specialized task instructions

Keeping a copy of important Gem instructions would reduce the risk of losing customized workflows during the transition.

Why the Change Matters

The retirement of Gems is significant because it reflects the direction in which consumer AI assistants are evolving.

Early AI assistants were largely conversational. Users asked a question, received an answer and started a new conversation for another task.

Custom assistants changed that model by allowing users to establish persistent behavior.

The next stage is increasingly about AI agents and reusable skills that can perform tasks with less repeated prompting.

Google’s reported move toward Skills therefore fits into a larger industry trend.

Gemini Is Becoming More Agentic

Google has increasingly emphasized AI systems capable of reasoning, using tools and completing more complex tasks.

The company’s Gemini ecosystem now includes features such as Deep Research and Canvas, alongside integrations with Google services.

Google’s recent Gemini updates have also emphasized coding, reasoning and the ability to create and refine content. Its release notes describe expanded capabilities across coding, image understanding and interactive creation through Canvas.

Moving from Gems to Skills could help align user customization with this broader direction.

Instead of creating a chatbot persona, users could increasingly define what they want Gemini to accomplish.

From Personas to Workflows

The change could also reduce the emphasis on AI “personas.”

Personality-based customization became popular because it made AI assistants feel more useful and personal.

However, businesses and advanced users often need repeatable workflows rather than personalities.

A sales employee might want an assistant that turns customer notes into a structured CRM update.

A researcher might want a tool that extracts findings from documents.

A developer might want a workflow that reviews code and produces a standardized explanation.

These use cases are better represented as capabilities than as standalone chatbot personalities.

Implications for Businesses

The transition could be particularly relevant to business users.

Companies increasingly want employees to use AI for repetitive knowledge-work tasks.

Reusable Skills could potentially provide a more standardized way of defining those tasks.

For example, an organization could create a workflow for preparing weekly reports, analyzing customer feedback or summarizing internal documents.

The major benefit would be consistency.

Instead of every employee developing a different prompt, organizations could potentially define standardized AI workflows.

Google would need to provide appropriate enterprise controls and security features for this to become a significant business use case.

Competition With OpenAI and Anthropic

Google’s change comes as the major AI companies increasingly compete around agents and customizable AI workflows.

OpenAI has been developing custom GPTs, agent capabilities and tools designed to let ChatGPT perform more complex tasks.

Anthropic has also emphasized Claude’s ability to work with tools and perform extended tasks.

The competitive battleground is therefore moving beyond model intelligence.

AI companies increasingly need to make their systems useful for completing real-world work.

A reusable Skills system could become an important part of Google’s response.

What Happens to Existing Gems?

The most important question for current users is how much of an existing Gem can be automatically transferred.

If Google provides a direct migration tool, users may be able to move their custom configurations with limited effort.

If users must manually recreate Skills, the process could be more disruptive for people who have built complicated Gems.

Google may also introduce new functionality during the transition that is unavailable in the current Gems system.

Until official migration documentation is released, the exact process remains unclear.

Potential Benefits of Skills

If the transition is implemented effectively, Skills could offer several advantages.

More Structured Workflows

Skills could provide a clearer framework for recurring tasks than free-form custom instructions.

Greater Automation

The feature could potentially allow Gemini to perform multiple steps automatically.

Better Reusability

Users could apply the same capability across different conversations and contexts.

Easier Scaling

Organizations could potentially standardize AI-assisted processes across teams.

These benefits depend on the final design Google provides.

Potential Challenges

The migration also creates several risks.

The biggest immediate issue is user disruption.

People who have spent time building customized Gems may not want to recreate them manually.

Another challenge is compatibility.

If Skills operate differently from Gems, simply transferring instructions may not reproduce the same behavior.

There could also be a learning curve for users who understand Gems but are unfamiliar with the new system.

Google will therefore need to provide clear migration tools and documentation.

Impact on AI Developers

The transition could also create opportunities for developers and advanced Gemini users.

If Skills become shareable or extensible, developers could potentially create specialized workflows for specific industries and use cases.

This could help build a broader ecosystem around Gemini.

However, the extent to which third-party developers can create, distribute or monetize Skills remains unclear.

Those details could determine whether Skills become a major platform feature or simply a replacement for Gems.

Google’s Broader Gemini Strategy

The reported Gems retirement should be viewed within Google’s larger effort to make Gemini the central AI layer across its products.

Google is integrating Gemini into its ecosystem rather than treating it solely as a standalone chatbot.

That includes connections with productivity tools, search, Android and other Google services.

The company has also been rapidly updating the underlying Gemini models and capabilities.

Google’s public Gemini API documentation, for example, shows an active cycle of model releases and deprecations as the company moves users toward newer model generations.

The rapid pace of change means users can expect more transitions as Google consolidates older features into newer AI frameworks.

What Users Should Watch

Gemini users should watch for official announcements covering:

• The final Gems retirement date
• Automatic migration options
• Whether Gem instructions can be exported
• How Skills work
• Whether existing Gem conversations remain accessible
• Whether uploaded files and reference materials migrate
• New Skills creation tools
• Availability across free and paid Gemini plans
• Enterprise support

These details will determine how disruptive the transition becomes.

Industry Impact

The reported shift from Gems to Skills reflects a broader transformation in consumer AI.

The focus is moving from simple customization toward reusable capabilities and task automation.

For Google, this could help make Gemini more useful for complex workflows. For competitors, it reinforces the need to develop their own agent and customization ecosystems.

For users, the transition may initially create some inconvenience but could ultimately provide a more structured way to build AI-assisted workflows.

Looking Ahead

Google’s reported plan to retire Gems by October 20 and replace them with Skills points toward a broader evolution of Gemini from a customizable chatbot into a platform for repeatable AI tasks. Gems helped users create specialized versions of Gemini through persistent instructions, while Skills appear designed to fit a future in which users expect AI assistants to perform more structured and reusable workflows.

The key issue now is execution. Google will need to make the migration straightforward, preserve users’ existing work and clearly explain what Skills can do that Gems cannot. The company will also need to demonstrate that the new system offers meaningful advantages rather than simply changing the name of an existing customization feature. As Google, OpenAI and Anthropic increasingly compete around AI agents and workplace automation, the ability to turn natural-language instructions into reliable, repeatable tasks could become one of the most important features differentiating the next generation of AI assistants.