Your Gemini Gems are getting a powerful makeover this November

For months, Gems have served as a handy utility for Gemini users seeking to tailor the AI’s behavior for specific tasks without repeatedly entering background context. By acting as customized mini-assistants designed for distinct topics, projects, or professional roles, Gems allowed individuals to streamline repetitive workflows. However, Google has decided to retire the feature in favor of skills, a broader and more integrated system that the company has already begun deploying in advanced environments like Gemini Spark.

To understand the change, one must look closely at what skills bring to the table compared to their predecessor. Industry observers and users familiar with competing AI ecosystems will likely recognize the concept of skills, which typically bundle instructions, reference files, examples, and scripts into a cohesive package. In the context of Gemini Spark, skills act as operational guideposts that instruct the AI agent on how to execute complex tasks, including which specific tools to employ and how to format the final output.

The transition to skills is expected to dramatically improve the user experience within Gemini chat windows. Under the current Gems system, users must navigate to a dedicated Gems section of the interface to access and activate their custom AI configurations. The upcoming skills framework, by contrast, is designed to allow users to invoke a skill directly within any active chat window simply by using a command prompt, such as typing a slash followed by the skill name—a methodology popularized by platforms like Claude and ChatGPT.

Furthermore, the new architecture addresses one of the primary limitations of Gems: the inability to combine multiple custom setups in a single interaction. Within Gemini Spark and the upcoming skills rollout, users can leverage multiple skills simultaneously within a single prompt, allowing for much more sophisticated and multi-layered AI assistance. The flexibility doesn’t stop there, as users can also ask Gemini to generate new skills dynamically based on specific, real-time requirements.

Collaboration and complexity are also major beneficiaries of the shift. Skills are inherently shareable, enabling seamless distribution across teams and collaborative workspaces. For advanced users tackling highly intricate workflows, the architecture even supports invoking one skill from within another, creating a cascading web of specialized AI capabilities that far surpasses the isolated nature of traditional Gems.

What happens to your Gems?

With the retirement of Gems scheduled for November, a central question for the user base has been the fate of previously created custom prompts. Google has confirmed that individuals who have invested time in crafting multiple Gems will not lose their work. Starting on the implementation date of November 17, 2026, the system will automatically begin migrating existing user Gems into the new skills format, ensuring continuity for personalized workflows.

Despite assurances regarding data migration, a significant point of ambiguity remains concerning accessibility across different subscription tiers. Presently, Gems are available to all Gemini users regardless of whether they maintain a free account or a paid subscription. However, skills are currently restricted to Google AI Pro and Ultra subscribers who utilize personal Google accounts. At the time of the announcement, Google had not yet clarified how the automated migration process will handle Gems created by free-tier users, leaving some uncertainty about whether those users will retain access to their migrated configurations under the new system.

Official guidance and detailed documentation from Google are expected to arrive alongside the rollout, though support pages outlining the transition mechanics were not yet live at the time of the announcement. Industry analysts suggest that this major infrastructural update to Gemini’s customization features could coincide with the broader release of Gemini 4. Such a timing would align the introduction of advanced skills with overall performance and efficiency improvements designed to close any remaining capability gaps between Google’s ecosystem and its primary competitors in the generative AI space.