Official source for the latest news and updates from across Google and Alphabet.
Mountain View, CA
Joined September 2017
- Tweets3.2K
- Following479
- Followers133K
- Likes434
Pinned Tweet
We're rolling out Gemini 4 Carbon, our strongest model built for your hardest problems and sustained agentic workflows across knowledge work, web development, and cyber defense.
While Argon remains the go-to for most software engineering and knowledge work, Carbon is built for deeply complex, long-horizon tasks that push a frontier model's intelligence:
🗂️ Turn dense folders of PDFs, charts, and blueprints into clean spreadsheets and reports. Carbon cross-checks details across files and flags missing data instead of guessing.
💻 Build complete web apps and 3D worlds with less back-and-forth. Carbon checks both its code and what's on screen to fix errors as it builds, and coordinates a team of agents from one prompt with /swarm in @Antigravity.
🛡️ Find and fix critical security bugs faster. Carbon for cyber helps security teams find, test, and patch real-world vulnerabilities in hours instead of weeks.
Google DeepMind researchers guided AI agents to mathematically prove key security properties of the code underlying the Signal protocol, the technology that protects billions of private conversations in apps like Signal and Google Messages. This breakthrough discovery was made using AlphaProof Code, a novel agentic framework for formal verification at unprecedented depth that combines human expertise with AI agents.
This is the first time the actual production code of a widely deployed protocol like Signal's has been formally verified – by any method, with or without AI – and at this depth. goo.gle/4z9U76M
2
860
Now available on Pixel 9 and newer: Google Fi's new super network.
• Real-time automatic switching between two of America's fastest 5G networks, so you hit fewer dead zones wherever you go
• Seamless, encrypted connections to millions of Wi-Fi networks with Wi-Fi Auto Connect+ (now rolling out to select Samsung & Motorola phones)
• Instant account help and troubleshooting with Ask Fi, a new support agent powered by Google AI directly in the Fi app.
News from Google retweeted
Introducing the new Professional Agentic Architect certification! Proving your expertise in building production-ready agents just got easier.
Used to train thousands of our own FDEs and partners, this certification is now available to all technical professionals who architect, build, and deploy agents on Google Cloud.
Check out more in our blog → goo.gle/3WXNC9R
News from Google retweeted
The Gemini app is now available directly on Apple CarPlay. iOS users can have free-flowing conversations with Gemini Live right from their car’s display, hands-free.
Use Gemini in CarPlay to catch-up on what you have coming up in your schedule, talk through your inbox, or brainstorm out loud.
Tap the Gemini icon on your CarPlay display on your next drive to try it out 🚘
The Gemini Decisions API is now in preview in @GoogleAIStudio, bringing our new Gemini Decision 1 model to workflows that need fast, closed-set decisions.
Built for workflows that need a quick, reliable call rather than a generated text answer, it scores multimodal inputs against up to 1,000 questions in a single pass (150–250 ms end-to-end) at nearly 5x lower cost across the shared benchmark suite than Gemini 3.5 Flash-Lite with the Interactions API.
It’s already powering dynamic model routing in @Gemini_Notebook (up to 40% lower latency), rapid playable wireframes in @stitchbygoogle, and emoji selection in CC from @GoogleLabs.
News from Google retweeted
Today we’re launching the Gemini Decisions API, powered by our new Gemini Decision 1 model in @GoogleAIStudio.
Instead of generating text token by token, it’s built for closed-set choices scoring up to 1,000 questions across text, images, audio, video + PDFs in a single pass. A few details:
- 3 decision types: yes/no, multiple choice (up to 255 options) + rubric scoring
- $0.10 / 1M input tokens with $0 billed for output tokens (~5x cheaper than Flash-Lite)
- leads on several benchmark evals with lowest calibration error
Google teams are already building with it, below is a look at Stitch Rapid designing a full product flow in seconds (1x speed)
We gave testers in our Fairwind Program early access to Gemini 4 Carbon, our strongest model with frontier performance in cybersecurity capabilities and long-horizon agentic workflows.
Carbon is built for tasks that test the limits of a model’s intelligence like complex, multi-step coding and debugging and open-ended scientific research and discovery.
Here’s what early testers have to say ↓ 🧵:
ALT Blue quote card from Shanks Salwan, Lead Applied AI Engineer at Hebbia: "Gemini 4 Carbon is the strongest model we've evaluated for deep financial research, building cleaner, better-structured Excel models and producing sharper PowerPoint decks."
ALT Blue quote card from Audrey Liu, Senior AI Engineer at Balyasny Asset Management: "Our benchmark is designed to evaluate the work financial teams actually do. Carbon stood out on tasks involving complex context and structured deliverables, especially charts, code, and tables."
ALT Blue quote card from Thai Tran, AI Product Lead at Glean: "Gemini 4 Carbon finds the call summary, the buried email or the exact pipeline figure across workplace data, then turns it into targeted analysis. In our evaluations, it led every frontier model we tested on task completion."
Scientists and researchers can apply for a chance to get access to both Argon and Carbon via @Antigravity’s highest usage tier free of charge for 6 months (onboarded on a rolling basis), subject to eligibility, capacity, and selection: goo.gle/4BQZydv
Learn more about Carbon and how to start building with it: goo.gle/48KvfrC
2
691
News from Google retweeted
🚨 NEW LABS EXPERIMENT 🚨
Introducing Tabletop 🎲— our newest experiment built to get your crew from "we should totally do that" to "we’re actually doing this."
Drop links and files into one shared space, add your crew, and ask Tabletop for whatever you need to create custom spaces on the fly:
🗺️ Interactive road trip maps & dinner-decider wheels
🏆 Live pickleball brackets & merch voting boards
🪑 Drag-and-drop seating charts (so you don’t get stuck next to your boss)
Pull up a chair and join the waitlist (US, 18+): labs.google/tabletop
Here’s what’s new in this month’s Pixel Drop:
Gemini Intelligence: Now available on Pixel 10.
Pixel Tag: Our first finder tag, available to order tomorrow.
Create custom widgets: Personalize your homescreen with custom widgets, like creating a spending tracker by typing “Show me my monthly shopping spend.”
New custom icon styles: Personalize your home screen icons with even more looks.
More enhanced features: Including an easier search interface and upgraded Gemini call screening.
Coming soon to Pixel Watch: Step-by-step workouts and more.
Learn more about the new features and updates coming to Pixel this month. goo.gle/4ziFoqz
News from Google retweeted
Gemini Robotics 2.5 is here. 🤖
Our latest physical AI is able to learn hundreds of everyday skills - from folding laundry and serving pastries to autonomously sorting car parts on the factory floor.
Here’s a snapshot of its capabilities 🧵
Clinicians spend nearly 28 hours a week on paperwork and admin tasks – while administrative back-office complexity accounts for up to 35% of all U.S. healthcare spending.
Today @GoogleCloud is introducing Gemini for healthcare in preview, connecting scattered records and helping with admin work so care teams can spend more time with patients and members.
We're launching this preview with an extensive partner ecosystem and customers. Learn more. goo.gle/4AkuAc5
