Google releases new apps for Windows and MacOS

· Source: AI - Ars Technica · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Novice, short

Summary

Google has released new native desktop applications for Windows and macOS, expanding beyond its traditional web-based product offerings. On April 15, 2026, Google officially launched a Windows search app, previously in beta since September, which allows users to search both the web and local files/applications. This app, accessible via Alt + Space, supports screen context sharing and includes AI Overviews. Concurrently, Google introduced its first standalone Gemini app for macOS, built in Swift using Google Antigravity. This Mac app, invoked with Option + Space, provides full Gemini web features, including file uploads, notebooks, Deep Research, Canvas, and generative AI models. Both apps offer contextual search capabilities, though the Windows app requires Windows 10 or 11 and is English-only, while the Mac app supports all Gemini regions and languages.

Key takeaway

For AI Product Managers evaluating platform strategy, Google's move to native desktop apps for search and AI signals a commitment to deeper OS integration. You should consider how dedicated desktop experiences can enhance user engagement and contextual utility for your own AI products, especially for features requiring local file access or screen context. Evaluate direct distribution models versus app store listings based on your target audience and feature set.

Key insights

Google launched dedicated desktop apps for Windows search and macOS Gemini, enhancing native OS integration.

Principles

Method

Google developed the macOS Gemini app in Swift using Google Antigravity, delivering over 100 features in under 100 days, bypassing the App Store for direct distribution.

In practice

Topics

Best for: AI Product Manager, General Interest, Tech Journalist

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI - Ars Technica.