Google is killing off Gemini’s Gems in favor of ‘skills’

In a significant reorganization of its artificial intelligence product strategy, Google is planning to sunset its Gemini "Gems" feature, according to a report from TechCrunch writer Sarah Perez. The technology giant is moving away from the customized, task-specific agent framework that characterized Gems, opting instead to transition toward a new system designated as "skills." This strategic realignment, dated September 28, 2026, highlights the rapid pace of product iteration among major technology firms as they vie for dominance in the consumer and enterprise AI sectors.

The decision to retire Gems reflects a broader industry shift toward more cohesive, multi-functional AI systems. The TechCrunch report notes that Google's pivot occurs alongside the rising popularity of comprehensive, all-in-one artificial intelligence agents developed by rivals. Specifically, Meta's platforms, Muse and Instinct, have gained substantial traction by offering unified experiences rather than requiring users to construct or manage separate, niche tools. In response to this competitive pressure, Google is adjusting its path, abandoning the modular, user-created agent concept in favor of a more integrated capability model.

Under the previous model, Gemini's Gems allowed users to build and deploy highly specialized agents designed to handle singular, narrow tasks. While this offered a high degree of customization, it also placed the burden of configuration on the user. The shift toward "skills" suggests a desire to deliver pre-configured, highly adaptable capabilities that work seamlessly within a unified ecosystem. By moving in this direction, Google aims to simplify the user experience, matching the frictionless, all-in-one utility that has driven the adoption of competing platforms in the rapidly evolving marketplace.

For founders, corporate executives, and technology leaders within the Valor & Ventures Media community, Google's strategic pivot offers vital insights into the trajectory of enterprise software. The transition from fragmented, custom-built micro-agents to comprehensive, platform-level intelligence indicates that the market is prioritizing immediate utility and broad integration over user-driven customization. Decision-makers planning long-term AI integration strategies must carefully monitor these vendor shifts to ensure their internal workflows remain compatible with the next generation of unified digital systems, avoiding over-investment in soon-to-be-deprecated custom configurations.