Google finally launched a native Gemini app for Mac, accessible anywhere with Option+Space to summon a floating chat bubble. The app lets you share any window with Gemini for contextual help, upload files from Google Drive, and generate images with Nano Banana or videos with Veo. It requires macOS 15+ and comes free with usage limits, plus paid tiers at $7.99 (AI Plus) and $19.99 (AI Pro) monthly.
This puts Google in direct competition with OpenAI and Anthropic for desktop AI dominance, but they're playing catch-up. Both ChatGPT and Claude's Mac apps already offer computer control — actually performing tasks on your behalf rather than just answering questions. Google's approach feels more like an enhanced Spotlight search than a true AI assistant. The timing is telling: this launches just one day after Google made its Windows desktop search widely available, suggesting a coordinated push to establish desktop presence before falling further behind.
What's missing from the coverage is how this fits Google's broader AI strategy. After my reporting on Gemini's 3D simulations feeling like catch-up work, this Mac app continues that pattern — solid execution of features competitors already perfected months ago. The window sharing and contextual help are useful, but without computer control capabilities, Gemini remains a glorified search interface while rivals build actual AI agents.
For developers and AI users, this means more choice in desktop AI tools, but ChatGPT and Claude still offer more practical automation. If you're already in Google's ecosystem and want quick answers without switching windows, Gemini works. But if you want an AI that can actually do things on your computer, stick with the incumbents until Google catches up on agent capabilities.