---
title: "Gemini 3.5 Flash Introduces Built-in Computer Use Tool"
source: "Google DeepMind：Blog（RSS）"
source_tier: "t1"
url: "https://deepmind.google/blog/introducing-computer-use-in-gemini-3-5-flash/"
permalink: "https://feed.astralor.com/items/itm_0ce8c6adc247f334d031f701"
published: "2026-06-24T16:30:01.000Z"
score: 71
channel: "ai"
tags: ["Gemini 3.5 Flash","Google DeepMind","Agent Automation","Computer Use Tool"]
---

# Gemini 3.5 Flash Introduces Built-in Computer Use Tool

## At a glance

Google DeepMind has integrated a computer use tool into Gemini 3.5 Flash, enabling the model to understand screenshots, simulate mouse and keyboard actions, and perform agentic automation. The tool requires no extra setup and can handle web navigation, data entry, interface testing, marking a key leap from conversation to action.

## Key points

- Gemini 3.5 Flash adds a built-in computer use feature with no external dependencies.
- The model can parse UI screenshots and generate precise mouse/keyboard action commands.
- Supports automated multi-step tasks like form filling and cross-page information extraction.
- This capability suits RPA, testing, and office assistant scenarios.

## Full text

A look at the built-in computer use tool in Gemini 3.5 Flash.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_0ce8c6adc247f334d031f701