Google Is Reworking Gemini Features: What Skills, Tasks and the End of Gems Mean

Google is reshaping its Gemini AI assistant by replacing Gems features with Skills and moving its background agent, formerly known as Spark, into a task-focused interface.

The changes affect how users save instructions, combine workflows and delegate work to AI. For everyday users and businesses, understanding the new Google Gemini features is important to avoid confusion, preserve existing workflows and make better use of the tools as they roll out.

Also Read: Google Launches Gemini 4 Argon With a Strong Focus on Cybersecurity

Why is Google Changing its Gemini Features?

Google is moving Gemini beyond a conventional chatbot experience, where users ask questions and receive answers, towards an assistant that can follow reusable instructions and manage more complex work.

The shift centres on two changes.

• First, Skills are replacing Gems, Google’s customised AI assistants.
• Second, Spark has been renamed Task, giving users a more direct way to delegate work to Gemini.

The distinction is important.

Skills determine how Gemini should approach a task, while Tasks represent the work a user wants the assistant to complete. Schedules can specify when that work should happen.

For example, a marketing professional could use a Skill to apply a company’s writing guidelines, assign Gemini a Task to prepare a campaign report, and schedule recurring work where the feature supports it.

Google aims to make these capabilities work together without forcing users to switch between separate assistants or repeatedly explain their preferences.

Its September 2026 announcement confirmed the transition from Gems to Skills, while its Gemini help documentation explains how Tasks, schedules and Skills fit together. Google Workspace’s announcement provides details on the transition.

Also Read: Google Launches Ask Google Pay: Gemini AI Assistant to Help Users Track Spending and Save Smarter

What are Gemini Gems and Why are they Being Phased Out?

Gems are customised versions of Gemini that users can configure for particular purposes. They follow user-supplied instructions, such as a preferred writing style, a specific role, or rules for completing a task.

A student might create a Gem to explain difficult subjects in simple language. A journalist could configure one to draft news reports in British English. At the same time, a business owner might build another to help prepare his agendas.

The benefit is consistency. Instead of entering the same instructions every time, users can return to a customised assistant that already knows the requirements.

However, Gems generally operate as separate assistants. Switching between them can create friction when a task requires multiple instruction sets or involves different stages of work.

Google is transitioning to Skills to make customised instructions reusable within regular Gemini conversations. Users can combine different Skills in one chat, allowing Gemini to follow several sets of instructions while working on the same task.

Google is phasing in the transition rather than completing it for every account immediately. Google says that personal accounts will move away from Gems by November 2026. Google plans to roll out the final transition for business & enterprise accounts by March 2027. Education accounts are scheduled to transition later, no sooner than June 2027.

Also Read: Top LLM Models for B2B Use in 2026: GPT-5, Claude, Gemini, Llama & More Compared

What are Google Gemini Skills and How Do They Work?

Gemini Skills are reusable sets of custom instructions that guide how the AI responds or performs particular tasks. They can save users from repeating routine directions and help maintain a consistent approach across different conversations.

The main difference from Gems is that Skills are designed to work within a unified chat experience. Users can combine them, and Gemini can automatically recognise when an enabled Skill is relevant to a request.

Skills are also built around the Markdown-based SKILL.md format, which lets you move compatible instructions between supported platforms. However, Google warns that Skills created in the Gemini app do not automatically synchronise with Skills created in Workspace apps. Users who need the same instructions in both environments must set them up separately.

What Can Users Do with Skills?

This version adds many capabilities for users. Here is what gets added to your existing skills with new Google Gemini features:

• Save recurring instructions: Store preferred writing styles, formatting rules and task-specific guidance.
• Combine multiple Skills: Apply brand guidelines alongside a report template/research framework.
• Use instructions in regular chats: Invoke a Skill without opening another customised assistant.
• Reduce repetitive prompting: Reuse instructions across suitable tasks instead of typing them again.
• Import compatible instruction sets: Use supported SKILL.md files to bring Skills created elsewhere into Gemini.

There are limits. Google’s documentation says Skills do not yet work with every Gemini feature, including Deep Research and Canvas. Users should check compatibility before rebuilding a workflow that depends on those tools.

Also Read: Google Pushes Gemini Deeper Into Android Ahead of Apple’s AI Reset

Gemini Gems vs Skills: What is the Difference?

Feature	Gems	Skills
Main purpose	Create a customised AI assistant	Provide reusable instructions
How they are used	Traditionally through a dedicated assistant	Within regular Gemini chats
Combining instructions	Often involves switching between assistants	Multiple Skills can be combined
Automatic application	Depends on how the assistant is used	Gemini can apply relevant enabled Skills
Transition	Being phased out	Becoming the replacement system

One practical change is greater flexibility. Skills let users combine instructions for a particular job without maintaining a separate assistant for every requirement.

For businesses, this could also make it easier to standardise routine tasks. A team could use a shared Skill for brand language and another for preparing client reports, helping employees follow the same process.

Also Read: “Gemini is my lawyer now”: Bengaluru founder turns to AI after alleged road rage incident

What Happened to Google Gemini Spark?

Google introduced Gemini Spark as an AI agent designed to do more than answer individual questions. It can handle multi-step tasks on a user’s behalf, helping with research, digital organisation, and workflows involving supported apps and services.

Google has since renamed Spark to Task. Its Gemini help documentation says the change reflects how users described their needs: they wanted Gemini to help them finish tasks. The new interface places Chat and Task options beside the prompt box, while existing tasks and schedules remain accessible through the sidebar.

The new name is more descriptive, but it also changes how users encounter the feature. Instead of looking for a separate Spark option, users can select Task when they want Gemini to work towards a larger objective.

For example, a user planning a business trip could ask Gemini to research travel options, compare hotels and organise the findings. Another user might ask it to review information across supported sources and prepare a document.

These are different from asking Gemini a single question. The agent is expected to work through multiple steps towards a defined outcome.

However, Tasks are not a guarantee of error-free or unrestricted automation. The work depends on the connected apps, permissions, available capabilities and instructions provided. Users should review important outputs and pay attention to any confirmation requests before allowing consequential actions.

Gemini Tasks vs Skills vs Schedules: How are they different?

The terminology can be confusing because the three features work together rather than compete.

Google’s documentation describes them as different building blocks: a Task is the objective, a Skill provides reusable instructions, and a Schedule determines when work should run or what event should trigger it.

Feature	What it does	Example
Skills	Tells Gemini how to approach a task	Apply a company’s brand guidelines to a report
Tasks	Defines the work Gemini should complete	Research competitors and prepare a comparison
Schedules	Determines when work should run	Prepare a technology news briefing every morning

A practical example illustrates how they fit together. A business could create a Skill that specifies how to format an internal report, assign Gemini a Task to compile weekly industry developments & use a Schedule to trigger the work at a set time.

The user still needs to check the information, confirm that the sources are appropriate and review the resulting report. Automation reduces repetitive work; it does not remove the need for oversight.

Tasks can also involve event-based triggers, such as responding to a flight delay, where supported. This differs from older Scheduled Actions in Gemini chat, which are designed around recurring or time-specific prompts. The two scheduling systems aren’t interchangeable.

Also Read: Google’s Gemini App Now Lets You Create Music From Text, Photos and Videos

What Should Gemini Users Do Before Gems Disappear?

Users who rely on Gems should prepare for the transition instead of waiting until the feature is removed from their account.

Google says remaining Gems will be automatically migrated to draft Skills when the relevant transition takes place. Users can also recreate them manually. Because the migration is phased, the timing depends on whether the account is personal, business, enterprise or education-based.

The following steps can help reduce disruption:

• Review existing Gems: Identify which assistants are used regularly and which contain important instructions.
• Save essential instructions: Keep a copy of the prompts and rules that underpin important workflows.
• Download knowledge files if migrating manually: Google provides instructions for saving files attached to Gems and adding them to Skills.
• Test the new Skills: Check whether the output follows the original requirements before relying on it for important work.
• Check feature compatibility: Some tools available through Gems may not yet work with Skills.
• Review account-specific dates: Business, enterprise and education users should consult Google’s official transition guidance.

Users should also remember that a Skill created in the Gemini app will not automatically appear in Workspace apps. If an organisation depends on the same instructions in both environments, it will need to recreate them separately.

The transition may be relatively straightforward for someone who uses a Gem to rewrite emails. A business with several customised assistants, supporting files and automated workflows will need a more careful review.

Also Read: Google Rolls Out Gemini AI-Powered “Auto Browse” in Chrome

How Could the Changes Affect Businesses and everyday Users?

For individual users, Skills could make Gemini more consistent and reduce the need to repeat instructions. A student might reuse a Skill to turn notes into revision questions, while a professional could apply a preferred format to reports and presentations.

Businesses could use Skills to preserve institutional knowledge. Instead of relying on employees to remember every formatting rule or process, a team could turn those requirements into reusable instructions.

Tasks introduce a separate opportunity: delegating multi-step work to an AI agent. For example, a research team could ask Gemini to gather information from supported sources, organise the findings, and prepare a document for review.

But there are certain risks associated. An AI agent working across connected services may encounter sensitive information or make changes that require human approval. Organisations should maintain access controls, review permissions and define which actions require sign-off.

The changes further create a learning curve. Users must understand which instructions belong in a Skill, which objectives should be assigned as Tasks & when a Schedule is appropriate. Clearer workflows will matter more than simply adding new names to the interface.

What Google’s Gemini Changes Mean for the Future of AI Assistants

Google’s shift from Gems to Skills and from Spark to Tasks signals a broader shift in AI development. Assistants are increasingly created to remember reusable instructions & help complete work, rather than respond to individual prompts.

For users, the benefit could be less repetitive prompting and more consistent results. For businesses, Skills, Tasks & Schedules could make it easier to standardise routine processes and delegate parts of a workflow.

The challenge is making these tools predictable, secure and easy to understand. Users need to know what the assistant can do, which permissions it requires and when human review is necessary.

For now, Gemini users should review their existing Gems, test Skills where available and understand the difference between assigning a Task and scheduling work. Google’s phased transition will determine when the old tools disappear from each account type.

The bigger change isn’t just a new set of names. Google aims to make Gemini a more integrated assistant that can follow instructions and help users complete work.