- Solan Sync
- Posts
- Gemini Nano AI for Chrome: Powerful Real-Time Scam Detection
Gemini Nano AI for Chrome: Powerful Real-Time Scam Detection
The on-device AI “Gemini Nano” now powers Chrome’s Enhanced Protection, detecting phishing, tech-support scams, and deceptive push notifications in real time. Early tests show an 80% reduction in scam results and block hundreds of millions of threats daily.
1. What Is Gemini Nano?
1.1 Benefits of On-Device AI
Privacy Preserved: Browsing data stays on your device—no cloud upload
Instant Response: No server round-trips for speedy scam detection
Offline Resilience: Works even on unstable networks
1.2 Integration with Chrome’s Enhanced Protection
Enable “Enhanced Protection” in Chrome settings and Gemini Nano activates automatically on both desktop and Android. No extra setup required to start on-device threat defense.

2. How Real-Time Scam Detection Works
2.1 Handling Phishing & Tech-Support Scams
Gemini Nano analyzes page structure and script behavior to spot:
Fake login screens
Suspicious dialogs
Unwanted redirects
When a threat is identified, Chrome shows a “Possible scam” warning with options to block, ignore, or learn why it flagged the page.
2.2 “Possible scam” Warning & User Actions
Block: Close the page for safety
Ignore: Proceed at your own risk
Details: View reasons behind the scam alert
3. Developer Cost-Saving Features
We just shipped implicit caching in the Gemini API, automatically enabling a 75% cost savings with the Gemini 2.5 models when your request hits a cache 🚢
We also lowered the min token required to hit caches to 1K on 2.5 Flash and 2K on 2.5 Pro!
— Logan Kilpatrick (@OfficialLoganK)
4:55 PM • May 8, 2025
3.1 Implicit Caching for 75% Token Discounts
The Gemini 2.5 API’s Implicit Caching automatically treats requests sharing a prefix as cache hits, cutting token usage by 75%—no configuration needed.
3.2 Reduced Prompt Lengths (Flash/Pro Models)
Flash Model: Minimum 1,024 tokens
Pro Model: Minimum 2,048 tokens
Shorter prompts mean lower costs, and usage metadata now reports cache-hit counts for better resource monitoring.
4. Key Features of the iPad Gemini App
updated image generation model is now live! time to experiment with gemini-2.0-flash-preview-image-generation🪄
notebook: github.com/google-gemini/…
— Nikita Namjoshi (@nikitanamjoshi)
5:06 PM • May 7, 2025
4.1 Split-View Multitasking
Run Gemini alongside other apps in Split View to streamline research, writing, and browsing.
4.2 Enhanced Text & Image Generation
Text Generation: More accurate long-form writing and summarization
Image Generation: Sharper text rendering and relaxed content restrictions
4.3 Deep Research & Canvas Integration
Deep Research: Guided research workflows
Canvas: Collaborative document and code editing
Ideal for teams creating reports, presentations, or code together.
5. Conclusion & What’s Next
For Users: On-device detection safeguards privacy and delivers instant protection.
For Developers: Implicit caching and shorter prompts slash costs.
For iPad Users: Multitasking and collaboration features boost productivity.
Expect ongoing updates to Chrome’s Enhanced Protection and Gemini APIs, further enhancing scam defense and developer support.
Reply