Tools #Google#AI agents Oct 31, 2026 · Originally published Oct 30

Gemini API's Computer Use tool is generally available across the Interactions and GenerateContent APIs, with Gemini 3.8 Flash as the recommended model

The tool that lets a model see a screen and issue clicks and keystrokes has moved from preview to GA, so teams building browser or mobile agents can now evaluate it as a production product.

Primary source Gemini API changelog · ai.google.dev Read the original ↗

// Key points

  • The October 30 changelog entry says the Computer Use tool is generally available (GA) across the Interactions API and the GenerateContent API.
  • The docs list Gemini 4 Argon, Gemini 3.8 Flash (recommended), Gemini 3.5 Flash-Lite (low latency, lower cost) and Gemini 3 Flash Preview, covering browser, mobile and desktop environments.
  • The model returns UI actions such as clicks, typing and scrolling with an intent field; your client executes them and sends back a new screenshot, with coordinates normalized to 0–999.
  • Seven built-in safety policy categories require confirmation or block actions, such as financial transactions, communications, account creation and legal agreements, and responses carry a safety_decision; screenshot prompt-injection detection is opt-in and off by default.

Builder's takeGA means you can now talk about stability and SLAs, but the execution layer is still yours: get one fixed flow working with Playwright in a sandboxed container before wiring it into a product. The first thing I would do is turn on the prompt-injection check that ships off, and I would not casually use disabled_safety_policies to drop the confirmation policies; this fits last week's pattern of letting agents act while hard-coding the boundaries.

// Background · from #Google

Full timeline →
  1. Oct 29 Google DeepMind announces Gemini 4 Carbon: first on the Artificial Analysis Intelligence Index at 59, going to cyber defenders first at $2/$10 per million tokens
  2. Oct 29 Google releases the Gemini API CLI in beta: a zero-dependency command-line client for coding agents, scripts and CI
  3. Oct 28 Google DeepMind unveils Genie 4, a natively multimodal world model that generates audio too, as a research preview for trusted testers

Every Monday, the week in AI in one email

Every Monday morning: the 6 AI stories that mattered most in the past seven days, with my take, in your inbox.

Your email is only used to send the weekly and is never shared. Every email has a one-click unsubscribe link. See the latest issue →