Point your phone at a stack of documents, and the system handles the rest.
In the quiet work of turning paper into data, Google has introduced something that shifts the burden from human patience to machine intelligence. The company's Document Scanner in Google Drive now uses on-device AI to handle multi-page capture, blur correction, and duplicate removal — all without sending a single image to the cloud. It is a small but telling moment in the longer story of how ambient computing is absorbing the friction of daily life, one tedious task at a time.
- The old page-by-page scanning ritual — tap, check, tap again — is being replaced by a single sweeping gesture as AI takes over the sequencing and cleanup.
- Privacy-sensitive users gain rare reassurance: all image processing happens locally on the device, with internet required only at the moment of saving.
- The feature demands real hardware muscle — Android devices with at least 8GB of RAM — drawing a sharp line between who gets the upgrade and who waits.
- Google is seeding the same capability into Files by Google and other Android apps, signaling this is infrastructure, not a one-app novelty.
- A gradual rollout means the experience will arrive unevenly, but the trajectory is fixed: document capture is becoming something phones simply handle.
Google has overhauled the Document Scanner inside its mobile Drive app, replacing a manual, frame-by-frame process with something far more fluid. Demonstrated by Android Ecosystem President Sameer Samat, the upgrade lets users hold their phone over a stack of pages and let the AI handle separation, correction, and cleanup automatically.
Three capabilities define the leap. The system identifies the sharpest frame in a sequence and discards blurry captures — sparing users the frustration of discovering a ruined scan after the fact. It also recognizes when the same page has been captured twice and silently removes the duplicate. Most notably, all of this happens entirely on the device; no data leaves the phone during scanning or image enhancement, only when the finished file is uploaded to Drive.
The trade-off is hardware. The feature requires Android devices with at least 8GB of RAM, a threshold Google set deliberately given the processing demands of real-time multi-page analysis. Google Play services handles the computation locally rather than routing it through the cloud.
The ambition extends beyond Drive. The same technology is being woven into Files by Google and other Android apps, framing it as a platform-level capability rather than a single product feature. The rollout is gradual, but the direction is clear: Google is treating smarter document capture as foundational to how people use their phones for everyday tasks.
Google is quietly reshaping how millions of people turn paper into digital files. The company has overhauled its Document Scanner feature in the mobile Google Drive app, moving from a manual, frame-by-frame process to something that feels almost like magic: point your phone at a stack of documents, and the system handles the rest.
The upgrade, demonstrated by Sameer Samat, Google's President of Android Ecosystem, represents a fundamental shift in how the scanning tool works. Instead of tapping to capture each page individually, users can now hold their camera over multiple sheets and let the AI do the heavy lifting. For larger documents, a simple pass of the phone across the page triggers automatic separation into individual scans. The interface itself has been redesigned with what Google calls "Material 3 Expressive" styling—cleaner, less cluttered, more professional.
Three specific capabilities make this feel genuinely useful rather than incremental. First, the system automatically detects and corrects for blur. Anyone who has ever scanned a document with shaky hands knows the frustration of getting a useless image and having to try again. Now the AI watches for the sharpest frame in a sequence and uses that, discarding the blurry ones. Second, if you accidentally capture the same page twice, the system recognizes the duplicate and removes it without asking. Third, and perhaps most important for privacy-conscious users, all of this processing happens on the device itself. No internet connection required for the actual scanning and image enhancement—only when you're ready to save the file to Google Drive.
The catch is hardware. The feature is rolling out on Android devices with at least 8GB of RAM. Google was explicit about why: real-time multi-page scanning, blur detection, and duplicate comparison demand serious processing power. The company chose to handle everything through Google Play services, keeping the work local rather than shipping data to the cloud.
This is not a Google Drive exclusive. The same capabilities are being built into other Android apps, including Files by Google, suggesting the company sees this as a foundational technology for the entire ecosystem rather than a single-product feature. The rollout is gradual, which means not everyone will see it immediately, but the direction is clear: Google is betting that making document capture faster and smarter will deepen how people use their phones for everyday tasks.
Notable Quotes
The substantial processing power and memory required for real-time multi-page scanning, blur detection, and duplicate comparison made this hardware restriction necessary— Google
The Hearth Conversation Another angle on the story
Why does Google care about making document scanning better? Isn't that a solved problem?
It's solved if you have patience and steady hands. But most people don't. The old way meant scanning page by page, dealing with blurry shots, manually deleting duplicates. It's friction. Google sees that friction as an opportunity to make the phone more useful for work.
The 8GB RAM requirement seems like a significant limitation. Who does that exclude?
Older phones, budget devices, anything below the mid-range. It's a real constraint, but it also tells you something: Google isn't compromising on the experience to reach everyone. They're saying this needs real horsepower to work smoothly.
Why keep it all on-device? Why not use cloud processing?
Privacy, mostly. And speed. If every blur detection and duplicate check required a round trip to Google's servers, the experience would feel sluggish. On-device means instant feedback. It also means your documents never leave your phone until you decide to save them.
Does this change how people will actually use their phones?
Probably not dramatically. But it removes a small daily annoyance for millions of people. That compounds. The person who scans receipts, contracts, or handwritten notes suddenly has a tool that doesn't fight them.