Google isn’t slowing down. Not after April, not even after rolling out the Gemini app itself. During Google I/O 2025 on Tuesday morning, they confirmed something specific: the macOS app gets an upgrade. Voice capabilities are coming. So is Gemini Spark, that new claw-like autonomous assistant everyone is talking about. All by summer.
Speaking your tasks
Josh Woodward took the stage. VP of Google Labs and head of the Gemini app team, he spoke from Mountain View. He had a MacBook. He had documents.
The demo was simple enough. Upload paperwork to the Finder. Woodward used his own life data. Vaccination records. Allergy lists. His pets, not tax returns.
Here’s the trick. Long-press the function key. Speak.
Don’t type. Don’t click through menus. Just tell it what to do. Woodward dictated an email instruction. Asked for a “friendly” tone. In the same breath, he told Gemini to convert the selected files into a table.
Release the key.
Gemini handles the rest.
“Because I’ve selected those files… its multimodal understanding kicks in,” Woodward said.
It reads the PDF. It looks at invoice images. Voice controls it all. Complex data turns into a structured table, right there. It works. It feels almost too easy, doesn’t it?
How to use it right now
The app exists. You can grab it at gemini.google/mac.
Launch it with a shortcut. Option plus Space. If you own a MacBook, it’s there. Waiting.
You already know the basics. Nano Banana image generation is baked in. That’s been live for a bit.
This isn’t new territory for phones. We’re used to chatting with bots on our wrists or screens. Native desktop integration is different, though. Heavier. More integrated.
Google wants your workflow on Apple hardware to feel… connected. It makes sense, considering Gemini is rumored to be the brain behind Apple’s redesigned Siri soon enough. The ecosystems are blurring. Lines getting fuzzy.
So, mark your calendars.
Summer is coming. The voice features drop then. Gemini Spark follows close behind.
