Google 发布 Gemini 4 Argon 前瞻模型并开放多端访问

AI 导读

Google 宣布推出最新前瞻模型 Gemini 4 Argon，面向 Google AI Ultra 和 Pro 订阅用户在 Gemini App 与 Search AI Mode 中开放，Antigravity 的 Ultra 订阅用户以及 Google AI Studio 和 Gemini 平台的开发者与企业可通过 API 使用。

推荐理由

原文列出 Argon 在 Chrome 代码重构、工程提交和机器人仿真上的具体落地数据，可据此评估前沿模型的实际工程价值。

正文 · 原文

Today we're rolling out access to Gemini 4 Argon, our latest frontier model built to solve your hardest problems.

Starting today, it’s available globally for Google AI Ultra and Pro subscribers in @GeminiApp and AI Mode in Search, Ultra subscribers in @Antigravity, and for developers and enterprises via the API in @GoogleAIStudio and the Gemini platform.

At @Google, Argon is already accelerating progress across engineering and robotics:

💻 @GoogleChrome: Argon helped the Chrome team refactor a Chromium 6-million-line legacy codebase, completing months of work in just weeks. For Chrome’s billions of global users, this eliminated an entire category of browser crashes, unlocking the ability for live webpages to move fluidly between tabs, side panels, and floating windows without reloading.

🚀 Software engineering: Googlers are using Argon to ship code faster than ever. In @Antigravity, they increased agentic code submissions by over 35% with rollbacks cut by nearly a third.

🤖 Robotics at @GoogleDeepMind: By watching video of a human playing cornhole, Argon autonomously built simulation code that taught it how to play, and helped Apptronik's Apollo humanoid sink seven straight shots.