# Google Launches General Office Agent Gemini Agent

- Category: Flash (快讯)
- Published: 2026-10-09T11:58:00.000Z
- Source date: 2026-10-09
- HTML: https://www.88z.ai/en/news/flash-20261009-e16165debf7c4bd6
- Markdown: https://www.88z.ai/en/news/flash-20261009-e16165debf7c4bd6.md
- Other language: https://www.88z.ai/news/flash-20261009-e16165debf7c4bd6.md

Google has launched the general office agent Gemini Agent, featuring autonomous planning and cross-application collaboration capabilities. This agent integrates into teams with an independent identity and optimizes task execution through four memory mechanisms, automatically selecting underlying models based on requirements.

## At a glance

- Gemini Agent can autonomously complete office tasks such as research, email writing, and presentation creation across applications
- Features four memory mechanisms and integrates into team collaboration as an independent AI colleague

## The update

Google has officially launched the general office agent Gemini Agent, designed to enhance workplace efficiency. This agent can autonomously plan complex tasks and invoke tools across different applications, such as Gmail, Sheets, and Slides, to independently complete work like information retrieval, email writing, and presentation creation. To integrate more naturally into team collaboration, Gemini Agent can create an "AI colleague" identity for each agent, complete with an independent email address and calendar. In terms of technical architecture, it is equipped with four memory mechanisms: conversational, semantic, procedural, and contextual, ensuring coherence in context understanding. Furthermore, the system can automatically select the most suitable underlying model for processing based on specific task requirements, achieving more efficient task delivery.

## Sources

- [量子位](https://mp.weixin.qq.com/s/ZTqTRMwppujjU2CbCi9JtQ) (2026-10-09)

This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.