Google Retires Gemini Gems as It Rebrands Custom AI Assistants as Skills

Google is retiring Gemini’s “Gems” and replacing them with a new system called “skills,” a move that will automatically carry users’ custom assistants into the updated format on November 17, 2026. The change matters because it signals how quickly Google is reshaping its consumer AI product as rival all-in-one assistants gain ground.

The company is notifying users inside the Gemini app that their existing Gems will be converted into skills without any action required. Until the transition date, the current feature will continue to work as normal. After that, the same personalized instructions and workflows will live under Google’s new naming and product structure.

What Google is changing in Gemini

Google is not deleting users’ work so much as renaming and reorganizing it. Gems, which let people create custom AI helpers for specific tasks, are being folded into a broader “skills” framework designed to be used across different kinds of AI interactions.

The update is the latest sign that Google is trying to simplify its Gemini app while also making it more adaptable. The company has increasingly leaned into modular AI features, but the shift also highlights a familiar Google pattern: introducing branded tools, then later consolidating them as the product evolves.

How do Gems become skills?

Google says the conversion will happen automatically. Users will not need to rebuild their assistants or manually export prompts, because the company plans to migrate existing Gems into the new format for them.

The transition notice appears in Gemini itself and sets November 17, 2026, as the date when Gems will officially become skills. Until then, people can continue using the old feature as usual.

Why Google built Gems in the first place

Gems were introduced in 2024 as a way to let people teach Gemini how to perform a task once and reuse that setup again and again. Instead of re-entering instructions every time, users could save a custom assistant tuned to a specific job.

The feature came with Google-made examples such as a learning coach, brainstorming partner, career guide, coding helper, and editor. Users could also build their own versions for daily life, including things like fitness coaching, meal planning, and travel planning.

Google also allowed those custom assistants to be shared, which suggested a broader ambition: Gems were not just a productivity feature but a potential growth tool for Gemini itself. Shared custom assistants could help demonstrate the app’s flexibility and attract more users into Google’s AI ecosystem.

What made Gems useful?

Gems were useful because they reduced repetition. A user could create a tailored assistant once, then return to it whenever the same type of task came up.

That made the feature appealing for people who wanted an AI app that felt more personalized than a general chatbot, while still being simple enough for everyday use.

• They let users save custom instructions for repeat tasks.
• They offered prebuilt assistants for common needs.
• They could be shared with other users.
• They gave Gemini a more customizable, product-like feel.

Why is Google replacing Gems now?

Google appears to be responding to a broader shift in the AI market, where companies are moving toward more unified assistant experiences rather than a menu of separate branded features. As all-in-one agents become more prominent, older product layers can start to look cluttered or redundant.

That shift is not happening in a vacuum. The source notes that competing systems such as Meta’s Muse and Instinct are helping push the market toward AI tools that feel more like full assistants and less like collections of isolated add-ons.

By replacing Gems with skills, Google may be trying to make Gemini’s feature set more flexible behind the scenes, even if the user experience remains somewhat technical.

Google’s decision reflects a wider industry move toward integrated AI assistants, where reusable capabilities are built into the assistant itself rather than presented as a standalone branded feature.

What does this say about Google’s product strategy?

It suggests that Google is still refining how it wants people to think about Gemini. The company has often introduced new tools with distinct names, icons, and app placements, only to reorganize them later when the product direction changes.

That approach can create confusion, especially for users trying to keep track of what belongs where. But it also reveals an effort to iterate quickly in a fast-moving AI category where product design is still being invented in public.

The critique is not unique to Gemini. Google has long been known for launching overlapping products and then consolidating them later, including in its messaging and communications businesses. The AI era, it seems, has not changed that habit so much as given it a new stage.

How skills will work inside Gemini

Skills are meant to carry forward the core idea behind Gems, but they are likely to be used in a more workflow-oriented way. Google says users will need to type a forward slash in a task thread to choose the skill they want to invoke, which makes the feature feel more like a command system than a plain-language chatbot conversation.

That detail matters because it exposes a tension in Google’s design. The company wants Gemini to feel smart and accessible, but the new interface sounds more like something built for power users or developers than for casual consumers.

In practice, that may make skills more capable for organized work, but less intuitive for people who simply want to ask a bot for help without thinking about commands.

Why does the slash command matter?

The slash command matters because it shows how Google is balancing flexibility with usability. A structured command system can be efficient for experienced users, but it adds another layer of interaction that may discourage people looking for a simple chat experience.

That is especially notable at a time when competing products are trying to reduce friction and make AI feel more conversational, less configured.

Timeline: From launch to shutdown

Gems have had a short but eventful lifespan. They were introduced in 2024 and are now being phased out roughly two years later, as Google aligns Gemini with a new organizational model.

How does this compare with rival AI assistants?

Google’s move comes as the market shifts toward more integrated agent-style products, where the assistant does not just answer questions but also carries out recurring workflows. In that environment, a separate branded feature like Gems can start to feel less important than the underlying capability.

That is one reason the update is strategically important. Google is not just changing terminology; it is trying to align Gemini with the way the next generation of AI tools may be organized.

Still, the new system may not instantly solve the usability challenge. If the experience remains too technical, users could continue to prefer simpler chatbot interfaces that require less setup.

What is the risk for Google?

The risk is that a cleanup intended to improve Gemini could instead make the product feel less approachable. Renaming Gems as skills may help Google streamline the backend, but it could also remove a bit of the personality and clarity that a branded feature provides.

For consumer AI products, memorability matters. A good name helps users understand what a feature does and where to find it. Replacing a recognizable concept with a more generic one can make a product more coherent to engineers while making it less friendly to ordinary people.

What users should know now

For current Gemini users, the practical message is simple: nothing needs to be rebuilt right away. Google says existing Gems will keep working until the migration date, and the company will handle the move into skills automatically.

If you have already spent time setting up a custom assistant, that work should not disappear. Google’s goal is to preserve those configurations while changing the framework behind them.

The main difference users may notice is how the feature is presented and accessed. As the update rolls out, Gemini is likely to feel less like a collection of separate tools and more like one assistant with layered capabilities.

• Existing Gems remain available until November 17, 2026.
• Google will migrate them automatically.
• Users should not need to recreate their assistants.
• The new system will be called skills.

The bigger picture for Gemini

On the surface, this is a product rename. In reality, it is a signal that Google is still in the middle of defining what Gemini should be. The company has big ambitions for its AI platform, but those ambitions require a user experience that is both powerful and easy to understand.

Moving Gems into skills may help Google standardize its AI architecture for the next phase of development. It may also reduce the confusion of having too many branded features competing for attention inside the app. But the change also underscores the challenge of building a consumer AI product that feels elegant rather than stitched together.

That challenge will only grow as the field shifts toward more agentic systems. If all-in-one AI assistants become the norm, the winning products will likely be the ones that can hide complexity without making users feel like they need a manual.

For now, Google is betting that skills will be a cleaner foundation for Gemini’s next chapter. Whether users experience that as simplification or another layer of complexity will depend on how well the company executes the transition.