Google's September AI recap: Gemini 4 Argon, Connected Apps and Googlebook

Gemini 4 Argon starts with a limited rollout

According to Google, "September's headline belongs to Gemini 4 Argon," its newest frontier model. The company says it was built for advanced reasoning and has an industry-leading 1-million-token output limit. Google lists coding, cybersecurity defense, financial research, legal drafting and autonomous cybersecurity patching among the workloads it is meant for.

Access is restricted for now. Google says releasing capabilities at this level requires a phased approach, so Argon is rolling out to trusted cyber defenders through the new Fairwind Program. The company says it is gathering feedback from early testers to iterate on guardrails before expanding availability to developers, enterprises and consumers. It did not give a date for that wider release.

Gemini 3.8 brings faster models, cyber tools and voice

Gemini 3.8 Flash is described by Google as its best reasoning and coding performance yet, at the same speed and low cost as 3.7 Flash. The company points to gains in long-horizon software engineering, multi-step problem solving and agentic workflows. Gemini 3.8 Flash Cyber targets defenders with autonomous vulnerability discovery and automated code patching, and is available to trusted organizations through the Fairwind Program.

On the voice side, Google launched two audio models: Gemini 3.8 Live for everyday conversation and Gemini 3.8 Live Extended Thinking, which it calls the top-rated speech-to-speech model for multi-step reasoning. Both are available to developers in Google AI Studio and the Gemini API, and to users across Search Live, Gemini Live and Google Workspace, including Gmail, Docs and Keep.

Google also introduced Gemini 3.8 Flash-Lite TTS and Gemini 3.8 Flash TTS, text-to-speech models that can generate custom voices and direct scene dialogue from text prompts. They are rolling out across products including Gemini Notebook and Google Vids.

New tools for everyday use

Gemini now supports Connected Apps from several categories. Productivity tools include Airtable, Linear, monday.com and PandaDoc. Creative platforms include Adobe, Picsart, Squarespace and Webflow. Lifestyle services include Peloton, SeatGeek and Experian. Users link them from their Gemini settings.

Lyria 3.5, which Google calls its most expressive music generation model yet, creates songs from text prompts or images. It is available globally in Gemini and across Google AI Studio, Google Flow Music and Google Vids.

Gemini Notebook gained coursework tools such as custom flashcards, quiz formats and 60-second Short Video Overviews in over 80 languages. Eligible college students can claim a free 12-month Google AI student plan with higher usage limits.

In Google Labs, CC is expanding from a personal agent into a group agent for household logistics and shared schedules. It gets its own email address to coordinate family calendars and track group to-dos, with built-in privacy controls. CC for groups is an experimental preview.

Googlebook, Windows and science projects

Googlebook is now available for pre-order. Google describes it as built on Android with desktop foundations from ChromeOS, designed for Gemini Intelligence and made to work with an Android phone out of the box. On-device features include Magic Pointer, which lets users point and ask Gemini about anything on screen with a quick cursor wiggle, and Rambler, a voice dictation tool that structures spoken notes. The Gemini app also came to Windows.

On the research side, Google highlighted AlphaGenome Atlas for mapping human DNA and the tracking of global methane from space. Project Suncatcher will launch its first test satellite with Planet to see how Google TPUs perform under real space conditions. Google says satellites in low-Earth orbit can access up to eight times more solar power than on Earth.

The company also said its tools support more than 300 languages spoken by 7 billion people, reaching 86% of the global population. WeatherNext 3, its flagship AI weather model, now includes real-time satellite data, hourly refreshes, higher resolution, precise precipitation forecasting and clean energy variables. Google added that Fast Company named it the 2026 Design Company of the Year.

What it means for users

For most people, the near-term changes are the Gemini 3.8 voice models, Connected Apps, Lyria 3.5 and the Notebook study tools, which Google says are already rolling out. Gemini 4 Argon and Gemini 3.8 Flash Cyber are different: both are limited to vetted security users for now, and Google has said broader access depends on feedback about guardrails. Developers can already use the Gemini 3.8 Live models through Google AI Studio and the Gemini API.